Astrophysics 2012

Page 1 of 249 Astrophysics_ASTR322_323_MCM_2012.
docx 29 Jun 2012

Stars, Galaxies, and Beyond
Summary of notes and materials related to University of Washington astronomy courses:
ASTR 322 The Contents of Our Galaxy (Winter 2012, Professor Paula Szkody=PXS) &
ASTR 323 Extragalactic Astronomy And Cosmology (Spring 2012, Professor eljko Ivezi=ZXI).
Summary by Michael C. McGoodwin=MCM.
Content last updated 6/29/2012

Rotated image of the Whirlpool Galaxy M51 (NGC 5194)
1
from Hubble Space Telescope HST,
with Companion Galaxy NGC 5195 (upper left), located in constellation Canes Venatici, January 2005.
Galaxy is at 9.6 Megaparsec (Mpc)= 31.3x10
6
ly, width 9.6 arcmin, area ~27 square kiloparsecs (kpc
2
)

1
NGC = New General Catalog, http://en.wikipedia.org/wiki/New_General_Catalogue
2
http://hubblesite.org/newscenter/archive/releases/2005/12/image/a/
Page 2 of 249 Astrophysics_ASTR322_323_MCM_2012.docx 29 Jun 2012

Table of Contents

Introduction ..................................................................................................................................................................... 3
Useful Symbols, Abbreviations and Web Links .................................................................................................................. 4
Basic Physical Quantities for the Sun and the Earth ........................................................................................................ 6
Basic Astronomical Terms, Concepts, and Tools (Chapter 1) ............................................................................................. 9
Distance Measures ....................................................................................................................................................... 9
Time Measures ........................................................................................................................................................... 10
Sky Position, Coordinate System, Motion and Rotation, Ecliptic, and Sidereal Terms .................................................. 12
Celestial Mechanics (Chapter 2) ...................................................................................................................................... 15
Keplers Original Laws of Planetary Motion .................................................................................................................. 15
Laws of Gravity and Motion......................................................................................................................................... 16
Light Properties (Pre-QM and Early QM); Distance and Magnitudes (Chapter 3) .............................................................. 18
Astronomical Distance, Luminosity, Magnitude ........................................................................................................... 18
Magnitude and Distance Modulus ............................................................................................................................... 19
Light (Photon) Wave vs. Particle Properties .................................................................................................................. 20
Color and Color Index ................................................................................................................................................. 25
The Special Theory of Relativity (Chapter 4, omitted) ...................................................................................................... 28
Quantum Properties of Light and Matter; Spectroscopy (Chapter 5) ................................................................................ 28
Spectroscopy .............................................................................................................................................................. 28
Light Interactions and Atomic Models ......................................................................................................................... 29
Quantum Mechanical Considerations ......................................................................................................................... 31
Telescopes (Chapter 6) .................................................................................................................................................... 33
Binary Systems and Stellar Parameters (Chapter 7) ........................................................................................................ 35
Mass Determination Using Visual Binaries.................................................................................................................. 35
Eclipsing Spectroscopic Binaries ................................................................................................................................. 36
The Classification of Stellar Spectra (Chapter 8) ............................................................................................................. 39
Harvard Classification................................................................................................................................................. 39
Statistical Mechanical Considerations ......................................................................................................................... 40
The Hertzsprung-Russell (H-R or HD) Diagram ........................................................................................................... 44
Stellar Atmospheres (Chapter 9, omitted)........................................................................................................................ 49
The Interiors of Stars (Chapter 10).................................................................................................................................. 50
Basic Equations .......................................................................................................................................................... 50
Energy Transport and Thermodynamics ...................................................................................................................... 55
The Main Sequence ..................................................................................................................................................... 57
The Sun (Chapter 11) ..................................................................................................................................................... 59
The Solar Interior ........................................................................................................................................................ 59
The Solar Atmosphere ................................................................................................................................................. 61
Miscellaneous Solar Topics ......................................................................................................................................... 66
Solar Cycle and Miscellaneous Variable Solar Activity ................................................................................................. 67
The Interstellar Medium; Protostar and Early Star Formation (Chapter 12) ..................................................................... 70
Interstellar Dust and Gas............................................................................................................................................ 70
Protostars ................................................................................................................................................................... 73
Pre-Main-Sequence Evolution, Young Stellar Objects, and the ZAMS .......................................................................... 78
Main Sequence and Post-Main-Sequence Stellar Evolution (Chapter 13) ......................................................................... 86
Evolution of the Main Sequence .................................................................................................................................. 86
Late Stages of Stellar Evolution (Post-Main Sequence) ................................................................................................. 90
Stellar Clusters ........................................................................................................................................................... 94
Stellar Pulsation (Chapter 14) ......................................................................................................................................... 97
Key Observations of Pulsating Stars ............................................................................................................................ 97
The Physics of Stellar Pulsation .................................................................................................................................104
Modeling Stellar Pulsation .........................................................................................................................................106
Nonradial Stellar Pulsation ........................................................................................................................................107
Helioseismology and Asteroseismology .......................................................................................................................109
The Fate of Massive Stars (Chapter 15) ..........................................................................................................................110
Post-Main Sequence Massive Stars That Evolve To Supernovae ..................................................................................110
Supernovae (Supernovas) ...........................................................................................................................................115
Gamma-Ray Bursts GRBs and Associated X-ray Emissions .......................................................................................122
Cosmic Rays and Solar Energetic Particles .................................................................................................................126
The Degenerate Remnants of Stars (Chapter 16) ............................................................................................................129
White Dwarfs (WDs) ...................................................................................................................................................129
The Physics of Degenerate Matter ..............................................................................................................................131
Neutron Stars (NS) .....................................................................................................................................................135
Pulsars ......................................................................................................................................................................138
General Relativity and Black Holes (Chapter 17, mostly omitted) ...................................................................................147
Close Binary Star Systems (Chapter 18) ........................................................................................................................148
Gravity Potentials in a Close Binary Star System........................................................................................................148
Circumstellar Accretion Disks ....................................................................................................................................151
A Survey of Interacting Binary Systems ......................................................................................................................153
White Dwarfs in Semidetached Binaries .....................................................................................................................156
Type Ia Supernovae ...................................................................................................................................................161
Neutron Stars and Black Holes in Binaries ................................................................................................................161
The Solar System & Planetary Systems (Chapters 19 through 23, omitted) ....................................................................173
The Milky Way Galaxy (Chapter 24) ...............................................................................................................................173
Milky Way Fundamentals, Morphology, and Coordinates............................................................................................176
Kinematics of the Milky Way ......................................................................................................................................183
Galactic Center ..........................................................................................................................................................187
The Nature of Galaxies (Chapter 25) ..............................................................................................................................189
Spiral and Irregular Galaxies .....................................................................................................................................190
Elliptical and SO Galaxies ..........................................................................................................................................197
Relative Numbers of Galaxies by Hubble Type ............................................................................................................200
Galactic Evolution (Chapter 26) .....................................................................................................................................203
Interaction of Galaxies ...............................................................................................................................................203
The Formation of Galaxies .........................................................................................................................................207
The Structure of the Universe (Chapter 27) ....................................................................................................................209
Summary of Distance Measuring Methods and The Extragalactic Distance Scale .......................................................209
The Expansion of the Universe ...................................................................................................................................215
Groups and Clusters of Galaxies ................................................................................................................................218
Active Galaxies (Chapter 28) ..........................................................................................................................................226
Classification and Characteristics of Active Galaxies ..................................................................................................226
A Unified Model of Active Galactic Nuclei (including Quasars) ....................................................................................237
Radio Lobes and Jets .................................................................................................................................................243
Using Quasars to Probe the Universe .........................................................................................................................244
Cosmology (Chapter 29, omitted) ...................................................................................................................................249
The Early Universe (Chapter 30, omitted) ......................................................................................................................249

Introduction

I have prepared this summary to assist in learning some of the materials relevant to the courses named.
These 300level (junior) college courses have provided a satisfying opportunity to take a more detailed but still
manageable look at astronomy and astrophysics applying to space well beyond the solar system. It is likely
that humankind will not reach other stars or galaxies in the foreseeable future, but our understanding of our
place in the universe is remarkably enriched by pursuing this knowledge. I thank professors Szkody and
Ivezi
3
for allowing me this opportunity to explore these many important topics in astrophysics.
Sources: The materials in this summary derive from the lecture notes, the assigned textbook, and many Web
and published scientific articles and other sources as noted. I do not consider Wikipedia to be a definitive
authority on any subject, but it is often quite useful as a starting point for more authoritative exploration, and
I have included a number of citations from it. In some cases, I have added emphasis or needed punctuation
to quotations. Diagrams used here are not directly copied from the textbook or from our lecture notes, but
have been obtained from the quoted primary or other Web sources. I have in most cases omitted derivations
of formulas and certain other relevant details such as most error ranges, but have tried to present the most
important conclusions in compact form, and always where to go to read the details. For images and
diagrams, I have where possible selected color examples, for clarity, didactic impact, and esthetics, but have
used monochrome sources where superior in content. Images are often shrunk but are still presented in full
resolution, so they can be viewed better by zooming in where desired.

3
English search terms: Zeljko Ivezic [ZXI]
Textbooks and Key Resources: The main textbook used in these courses in 2012 is An Introduction to
Modern Astrophysics, 2nd Edition, by Bradley W. Carroll, Dale A. Ostlie, 2007 (my copy is 6th printing 2011),
Pearson/Addison Wesley, hereafter abbreviated IMA2. References herein to chapters refer to that textbook,
but a great deal of information I have included under these chapter headings is from the lectures and
especially from other sources, so may not have been stated as such in the textbook (nor have I always kept
the order in which material is presented in the textbook). This outline summary should not be used in place
of buying this excellent textbook (which is purported to be for sophomore level astrophysics). The textbook
covers many more topics, is vastly more detailed in what it covers, and is well worth the high price. A 3rd
edition would be eagerly received and is needed to incorporate some of the exploding new astronomical
knowledge.
I also refer occasionally to Burke BF and Graham-Smith F, An Introduction to Radio Astronomy, 3rd Ed.,
Cambridge, 2010, hereafter abbreviated IRA3.
DOI (Digital Object Identifier) citations may be resolved if necessary at http://dx.doi.org/ . Enter the DOI
without the DOI:. For instance, the article with DOI: 10.1146/annurev-astro-082708-101642 may be
found at http://dx.doi.org/10.1146/annurev-astro-082708-101642.
Etymology and definitions in part derive from the Oxford English Dictionary, online version accessed online
JanuaryJune 2012, hereafter abbreviated OED.
I have generally favored MKS SI units (as does the textbook) rather than the CGS system of units preferred by
many astronomers.
Author: I am a retired physician and an auditing student, and claim no expertise in this field. I have
included material cited from many scholarly and technical articles, but I do not pretend to have mastered the
contents of these articles (though I enjoy looking through them, trying to learn what I can).
Copyrights: You may use this document for educational non-commercial purposes. Please be aware that it
contains some copyrighted material. If you incorporate this document or parts of it into a presentation or
website etc., please acknowledge my authorship of the document, the documents URL

http://www.mcgoodwin.net/pages/astrophysics_2012.pdf ,
as well as the external sources I have cited. This is a not-for-profit informal personal study aidif you
publish any of it you should consider securing permission to publish the materials from other authors that I
have included. If you are an author or copyright holder and object to anything I have included as beyond fair
use, please advise me and I will make appropriate corrections.
This file will not be updated beyond 2012, so gradual obsolescence is to be expected.
Constructive corrections and clarifications would always be appreciated. Send these to:
MCM at McGoodwin period NET (please convert to standard format when using)

Useful Symbols, Abbreviations and Web Links

Characters and symbols useful in writing this document
Greek 12:
Greek 10:
Math & Misc. 12: ^
Math & Misc. 10: ^

Abbreviations, Acronyms, and Definitions
Most abbreviations and acronyms are spelled out at first use, and definitions are similarly worked into the
text at the first or most substantive use of a term.

Astronomy Links from Professors Szkody and Ivezi, and Other Sources
These are links from Professor Szkody, Professor Ivezi, and otherwise. There are also many links on specific
topics given throughout the document.
= Webpages I found to be of greater interest; = the very best or most useful to me
All-Sky Milky Way Panorama (Axel Mellinger, can vary tilt)
American Astronomical Society AAS
American Association of Variable Star Observers AAVSO
Cepheid Variable Stars in LMC Site (see also here)
Chandra X-ray Observatory (CXO, 1999): one of NASAs 3 remaining Great Observatories in space
Constellations
Cosmic Calculator (converts redshifts to comoving distance, light travel times, etc.)
Eclipsing Binary Stars Luminosity Simulation (needs JAVA)
Elements Optical Absorption/Emission Lines (needs JAVA)
Galactic Center
Galactic-Scale Interactions (MPEG movies)
General Catalog of Variable Stars GCVS
Giant Magellan Telescope GMT (future ~2020, seven 8.4 m segments, Las Campanas Obs., Chile)
Guide Star Catalog GSC (14th mag)
http://heasarc.gsfc.nasa.gov/High Energy Astrophysics Science Archive Research Center HEASARC
(NASA missions)
Hipparcos Space Astrometry Mission (ESA, high precision star positions datasets)
James Webb Space Telescope JWST (future NASA mission ~2018, IR-optimized 6 m)
Large Synoptic Survey Telescope LSST (8.4 m, 3200 MP, 3.5 wide field telescope,
Cerro Pachn, Chile, first light 2019
4
)
NASA Today
NASA/IPAC Extragalactic Database (NED
http://www.noao.edu/National Optical Astronomy Observatory NRAO (by Assn. of Universities for
Research in Astronomy)
N-Body Shop Galaxy simulations (MPG movies and home page)
Orbiting Binary Stars Radial Velocity Simulation (needs JAVA, adjust M1, M2, a, e, i, w)
Planetary Nebula Spectra
Pulsar Sounds
Pulsation Mode in White Dwarf Simulation (needs JAVA)
SIMBAD (Set of IDs, Measurements, and Bibliography For Astronomical Data dataset,
CDS=Centre de Donnes astronomiques de Strasbourg, France; )
Sloan Digital Sky Survey SDSS (star dataset w 2.5 m telescope, Apache Point Obs., NM)
Smithsonian Astrophysical Observatory SAO
Solar and Heliospheric Observatory SOHO (including Current Solar Images)
Space Telescope Science Institute (including Hubble Space Telescope HST, 1990)
Space Telescope Science Institute (STScI) Digitized Sky Survey (DSS):
includes Digitized POSS (Palomar Observatory Sky Survey), Hubble HST Phase 2
Spitzer Space Telescope (2003): one of NASAs 3 remaining Great Observatories in space, IR
Rotating Sky Explorer (Flash animation)
United States Naval Observatory USNO Astronomical Applications (Almanac)
UW Astronomy Department
Space Weather (EarthSolar Interactions)
Spaceweather.com
NOAA Space Weather Prediction Center
NOAA National Geophysical Data Center page on Solar-Terrestrial physics

4
http://www.lsst.org/files/img/LSST_Timeline.jpg
Basic Physical Quantities for the Sun and the Earth

Symbols used are Sun = and Earth =
Radius of Sun R
Equatorial Radius
5
= 6.955 x 10
8
m = 695,500 km = 432,163 mi = 109 R
Mean Diameter: 1.392 x 10
6
km
Radius of Earth (RE or R)
6

Mean at Equator = 6378.1 km
Mean at Poles = 6356.8 km
Ellipticity of Sun f (Flattening or oblateness)
7
= 910
6
(thus the Sun is a nearly perfect sphere)
Ellipticity of Earth = 0.00335 = ~0.3%
Distance from Sun to Earth, average (1 Astronomical Unit = 1 AU)
8

1.49610
8
km = 1.49610
11
m = 92.96 x 10
6
mi = 4.848110
6
parsec = 215 R = 108 D
= 8.32 lm (light minutes) = 499 ls (light seconds)
Mass of Sun M
9
= 1.989 x 10
30
kg = 333,000 M
Mass of Earth M = 5.9736 x 10
24
kg
Density of Sun [note that water density = 1 gm/cm
3
= 1000 kg/m
3
]
Mean: 1.408 gm/cm
3
= 1408

kg/m
3
= 0.255 mean
Central or Core: up to 150

g/cm
3
= 162 x 10
3
kg/m
3
(the core extends to 0.2 to 0.25 R)

Photosphere
10
= 2x10
-7
g/cm
3 =
2x10
-4
kg/m
3
[~1.6x10
-4
of Earth sea level atmos. density
11
]
Particle density = ~10
23
particles m
3

Corona Mean
12
= 110
15
g/cm
3
= 110
12
kg/m
3

Lower Corona
13
= 1x10
-16
g/cm
3

Particle density
14
= 10
15
10
16
particles m
3

Density of Earth
15

Mean = 5.515 gm/cm
3
=

5515 kg/m
3

Inner Core = 12.8 13.1 gm/cm
3
=

1280 1310 kg/m
3

Mantle = 3.4 5.6 gm/cm
3
=

3400 5600 kg/m
3
Crust = 2.2 2.9 gm/cm
3
= 2200 2900 kg/m
3

Atmosphere (at sea level)
16
=1.2x10
-3
gm/cm
3
= 1.2 kg/m
3

Temperature of Sun
17

Surface effective black body temperature Teff,= 5777 K = 9,939 F (IMA2, by StefanBoltzmann law)
Central
18
or Core (modeled)= 15,700,000 K

5
Solar Radius:
http://solarscience.msfc.nasa.gov/ and
http://en.wikipedia.org/wiki/Sun
6
http://en.wikipedia.org/wiki/World_Geodetic_System and http://en.wikipedia.org/wiki/Earth
7
Ellipticity:
http://en.wikipedia.org/wiki/Flattening:
Ellipticity is defined for an ellipsoid with equatorial radius a and polar radius b as (a-b)/a
8
http://en.wikipedia.org/wiki/Astronomical_unit
9
http://nssdc.gsfc.nasa.gov/planetary/factsheet/sunfact.html
10
http://en.wikipedia.org/wiki/Photosphere
11
http://solarscience.msfc.nasa.gov/
12
13
http://solar-center.stanford.edu/vitalstats.html
14
15
http://en.wikipedia.org/wiki/Earth
16
http://en.wikipedia.org/wiki/Earth's_atmosphere
17
Luminosity of Sun L, Bolometric
19,20
= 3.846 x 10
26
W = 3.846 x 10
33
erg/s
Luminosity of Sun
(in the Blue-Band B)

21
1.9 x 10
26
W (note: rarely used in comparisons and ratios)
Mass / Luminosity Ratio of Sun
= 5172 kg/W (calculated with above values).

Intensity of the Sun, Mean I
22
= 2.00910
7
Wm
2
sr
1

Stellar Classification of Sun, Spectral and Luminosity (MorganKeenan)
23,24
: G2V
Magnitude of Sun:
25

Absolute Visual (MV at 10 pc): +4.83
Apparent Visual (V = mV): 26.74
Elemental Composition of Sun
Photosphere (by number % or ppm of nuclei)
26
= H 91.0%, He 8.9%; O 774 ppm, C 330 ppm,
Ne 112 ppm, N 102 ppm, Fe 43 ppm, Mg 35 ppm, Si 32 ppm, S 15 ppm
Photosphere (by relative mass)
27
= 70% H, 28% He, 2% metals (C, N, O, etc.)
Central (by relative mass)
28
= 35% H, 63% He, 2% metals (C, N, O, etc.)
Age of Sun
29
= 4.57x10
9
yr
Age of Earth
30
= 4.54x109 yr
Rate of Mass Conversion to Energy of Sun
31
= 4.3 x 10
9
kg s
-1

(i.e., the equivalent mass m=E/c
2
converted to energy to yield the Suns bolometric luminosity)
Sun Mean Energy Production per Unit Mass
32
= 0.194 x 10
-3
J/kg = 8.1x10
-7
kcal (human calories)/kg
Rotation Periods of the Surface of the Sun:
33

Sidereal equatorial: 24.47 d (rotation period in days to same location in star frame)
Synodic equatorial: 26.24 d (rotation period in days to same apparent location as viewed from Earth)
Sidereal Carrington (~26 degrees latitude): 25.38 d
Synodic Carrington (~26 degrees latitude): 27.2753 d

18
ibid.
19
ibid.
20
Luminosity:
http://en.wikipedia.org/wiki/Solar_luminosity:
Luminosity here is total bolometric (wide spectrum) photon radiant energy output, and does not include
neutrino radiant energy, which adds 0.1x10
26
W
21
www.astro.sunysb.edu/aevans/PHY523/.../useful-definitions-pp.pdf
22
23
Star Spectral classification:
http://en.wikipedia.org/wiki/Spectral_classification:
For the Sun classified as G2V in the MorganKeenan classification,
Letter G indicates a yellowish star with surface T 5,2006,000 K
Number 2 indicates two tenths of the range between star spectral class G0 and start of adjacent class K0
Roman Number V indicates the width of certain absorption lines, which correlates with star sizeV
indicates a Main Sequence dwarf star.
24
http://en.wikipedia.org/wiki/Hertzsprung%E2%80%93Russell_diagram
25
26
27
28
ibid.
29
http://en.wikipedia.org/wiki/Sun and http://solarscience.msfc.nasa.gov/
30
http://en.wikipedia.org/wiki/Age_of_the_Earth
31
32
ibid.
33
http://en.wikipedia.org/wiki/Solar_rotation
Solar Rotation period (an empiric formula by latitude at the surface for data collected 1967 1987):
34

= 14.713 2.396sin
2
() 1.787sin
4
() , where
= deg/d sidereal
= solar latitude (deg)
This gives sidereal days: 024.47 d; 2625.38 d; 3025.71 d; 6030.22 d; 9034.19 d
Total Solar Irradiance (TSI, aka the Solar Constant) is the irradiance (W m
-2
)

at all wavelengths
of photons (effectively from about 10,000 nm to about 10 nm) at exactly 1 A.U.
35

on a surface perpendicular to the incoming rays:
Mean is per NASA SORCE
36
about 1361 W m
-2
(was previously said to be 1366 W m
-2
)
Instantaneous solar irradiance
37
varies due to changing EarthSun distance
and solar fluctuations from 1,412 to 1,321 W m
-2
Obliquity or Tilt of Suns rotational axis
38
with respect to the ecliptic plane: ~7.25 degrees
Tilt of Suns magnetic dipole field axis
39
with respect to rotation axis: up to (10 to 20)
Sun rotation direction: Counterclockwise (when viewed from the north)
(this is the same direction that the planets including Earth rotate and orbit around the Sun).
Sun Escape Velocity: Escape velocity
40
in general is given by ve = (2GM/r)
1/2
, where M is the mass of the
massive body being escaped from starting at distance from the bodys center r, and ve is the minimum
speed required (ignoring drag) when propulsion ceases at that distance. For the Suns surface, the
idealized escape velocity is calculated at 617.5 km/s, whereas for the Earth surface, it is calculated at
11.2 km/s. However, there appears to be a complex relationship between this simplistic solar escape
velocity estimate and the actual velocities attained in solar wind. The latter is found to be below ve at least
using estimates from actual measurements near the Sun (at 4R 7R) using data from the Ulysses
probe,
41
and certainly solar wind speed is often < ve by the time the solar wind has decelerated in travelling
to Earth orbit.

34
Solar differential rotation:
Snodgrass, H. B. & Ulrich, R. K., Rotation of Doppler features in the solar photosphere,
Astrophysical Journal, Part 1 vol. 351, 1990, p. 309316
35
ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SOLAR_IRRADIANCE/COMPOSITE.v2.PDF
36
http://en.wikipedia.org/wiki/Sunlight
37
http://en.wikipedia.org/wiki/Solar_irradiation
38
http://solarscience.msfc.nasa.gov/sunturn.shtml
39
http://www.nso.edu/press/newsletter/Tilted_Solar_Magnetic_Dipole.pdf
40
http://en.wikipedia.org/wiki/Escape_velocity
41
Efimova AI et al. Solar wind velocity measurements near the sun using Ulysses radio amplitude
correlations at two frequencies. Advances in Space Research. Volume 35, Issue 12, 2005, Pages 21892194
Basic Astronomical Terms, Concepts, and Tools (Chapter 1)

This is only a limited listing of some of the topics and definitions in this chapter.
Distance Measures

These are given with various degrees of rounding.
(Image depicts typical distance measurement using geometric parallax.)
42

Astronomical Unit AU
The earthsun distance
43
= 1.4960x10
11
m 1.50x10
8
km 93 million miles
8 lightmin
Lightyear or light-year or light year ly
The distance light travels in 1 Julian yr = 9.46110
12
km = 9.46110
15
m.
computed by = 365.25 * 86400 s * 299,792,458 m/s =9.4607310
15
m
Parsec (distance at which 1 AU subtends 1 second of arc) = 206,265 AU = 3.26
ly = 3.110
13
km
Hubble Law, H0 and h
v = H0d where d = proper distance to a galaxy in Mpc, v = recession proper
velocity in km s
-1

H0 = 100h km s
-1
Mpc
-1
where the WMAP 2010 value of the dimensionless

Speed of Light Speed and Light Travel time/distances
c (speed of light) = 299,792,458 m/s (exactly, by definition) 3x10
8
m/s
3x10
5
km/s 186,282 mi/s
1 lightsec = 3x10
5
km
1 lightmin = 18 million km (3x10
5
km/s x 60s)
1 lightyear (ly, see above)
Representative distances
Earth Radius, equatorial mean= 6,377 km (diameter = 12,754 km)
Earth to moon distance, mean
44
384,400 km 239,000 mi 1.3 lightsec
Sun to Pluto distance = 30 to 49 AU, 5 lighthours
Sun to Oorts Cloud 5,000 to 50,000 AU
Sun to Proxima Centauri (nearest star) = 4.24 LY
Sun to Large Magellanic Cloud (a nearby galaxy
45
) 157 kly (48.5 kpc)
Milky Way disk diameter (visible) 50 kpc 163,000 LY
Milky WayAndromeda Galaxy 778 kpc

42
http://en.wikipedia.org/wiki/Parallax
43
http://en.wikipedia.org/wiki/Astronomical_unit
44
http://en.wikipedia.org/wiki/Lunar_distance_%28astronomy%29
45
http://en.wikipedia.org/wiki/Large_Magellanic_Cloud
Distance Measured
by Parallax
Time Measures

See further below for sidereal time and epochs
Astronomical Year Numbering
Astronomical year numbering is based on AD/CE year numbering, but follows normal decimal integer
numbering more strictly. Thus, it has a year 0, the years before that are designated with negative
numbers and the years after that are designated with positive numbers. Astronomers use the Julian
calendar for years before 1582, including this year 0, and the Gregorian calendar for years after 1582...
The year 1 BC/BCE is numbered 0, the year 2 BC is numbered 1, and in general the year n BC/BCE is
numbered "(n 1)". The numbers of AD/CE years are not changed and are written with either no sign or
a positive sign. The system is so named due to its use in astronomy... Although the absolute numerical
values of astronomical and historical years only differ by one before year 1, this difference is critical when
calculating astronomical events like eclipses or planetary conjunctions...
46

Gregorian Calendar
The Gregorian calendar, also called the Western calendar and the Christian calendar, is the
internationally accepted civil calendar. It was introduced by Pope Gregory XIII (1502 - 1585), after whom
the calendar was named, by a decree signed on 24 February 1582... The motivation for the Gregorian
reform was that the Julian calendar assumes that the time between vernal equinoxes is 365.25 days, when
in fact it is presently almost exactly 11 minutes shorter. The error between these values [resulted in an
accumulated error of about 10 days by the time of the reform]...
47
The rule for determining Leap Years
now requires that a Leap Year (1)be evenly divisible by 4; (2) not evenly divisible by 100 unless also evenly
divisible by 400. These rules have the effect of omitting 3 Leap Years every 400 years out of the 100 that
are divisible by 4. The average [Gregorian] year length is 365+(97/400) = 365.2425 days per year, a close
approximation to the tropical year of 365.2420 days(see below). The proleptic Gregorian calendar is
produced by extending the Gregorian calendar backward to dates preceding its official introduction in
1582.
48

Heliocentric Julian Day HJD
is the Julian Date (JD) corrected for differences in the Earth's position with respect to the Sun. Even more
precise corrections are made with the Barycentric Julian Date. Due to the finite speed of light, the time
an astronomical event is observed depends on the changing position of the observer in the Solar System.
Before multiple observations can be combined, they must be reduced to a common, fixed, reference
location. This correction also depends on the direction to the object or event being timed.
49
The BJD,
which may differ from the HJD by up to 4 seconds, has replaced the HJD for precise time comparisons.
Julian Calendar
The Julian calendar is a reform of the Roman calendar introduced by Julius Caesar in 46 BC (708 AUC).
It took effect the following year, 45 BC (709 AUC), and continued to be used as the civil calendar in some
countries into the 20th century. The calendar has a regular year of 365 days divided into 12 months.... A
leap day is added to February every four years. The Julian year is, therefore, on average 365.25 days
long.... The calendar year was intended to approximate the tropical (solar) year. Although Greek
astronomers had known, at least since Hipparchus, that the tropical year was a few minutes shorter than
365.25 days, the calendar did not compensate for this difference. As a result, the calendar year gained
about three days every four centuries compared to observed equinox times and the seasons. This
discrepancy was corrected by the Gregorian reform, introduced in 1582.
50
The proleptic Julian calendar is
produced by extending the Julian calendar backwards to dates preceding AD 4 when the quadrennial leap
year stabilized.
51

Julian Date JD
This is the number of days in the Julian Calendar since January 1, 4713 BCE at noon at Greenwich, i.e.,
by UT.

46
http://en.wikipedia.org/wiki/Astronomical_year_numbering
47
http://en.wikipedia.org/wiki/Gregorian_calendar
48
http://en.wikipedia.org/wiki/Proleptic_Gregorian_calendar
49
http://en.wikipedia.org/wiki/Barycentric_Julian_Date
50
http://en.wikipedia.org/wiki/Julian_calendar
51
http://en.wikipedia.org/wiki/Proleptic_Julian_calendar
JD with fractional value of .0 is at noon UT
JD with fractional value of .5 is at midnight UT.
(Jan 1, 2012 at 0h UT = 2,455,927.5 JD and Jan 1, 2012 at noon UT is 2455928.0 JD)
See here
52
for a JD calculator.
Modified Julian Day MJD
JD minus 2,400,000.5 day (used by spacecraft)
Note that with MJD, a fractional value of .0 is at midnight UT. One second after midnight, the MJD integer
remains the same.
Second (SI)
The standard SI second is the duration of 9,192,631,770 periods of the
radiation corresponding to the transition between the two hyperfine levels of
the ground state of the cesium 133 atom
53
There are 60x60x24=86,400 SI
seconds in the Julian day and 60x60x24x365.25=31,557,600 seconds in the
Julian year. No other type of second is in common scientific usage, despite
the confusion in days and years.
Terrestrial Time TT
a modern astronomical time standard defined by the International
Astronomical Union, primarily for time-measurements of astronomical
observations made from the surface of the Earth... For example, the
Astronomical Almanac uses TT for its tables of positions (ephemerides) of the
Sun, Moon and planets as seen from the Earth. In this role, TT continues Terrestrial Dynamical Time
(TDT),.. which in turn succeeded ephemeris time (ET). The unit of TT is the SI second, the definition of
which is currently based on the cesium atomic clock, but TT is not itself defined by atomic clocks. It is a
theoretical ideal, which real clocks can only approximate. TT is distinct from the time scale often used as a
basis for civil purposes, Coordinated Universal Time (UTC).
54
TT takes into account effects of special and
general relativity as Earth moves about the Sun and rotates. (IMA2 p. 15)
Universal Time UT or GMT, UT1, UTC
local time at Greenwich (UT=PST+8 hr = PDT+7 hr). UT1 is defined by Earths actual rotation and thus
based on irregular gravity effects. UTC is an acronym which stands for Universal Time Coordinated, a
linguistic compromise between English Coordinated Universal Time and French temps universal
coordonn. UTC is measured by atomic clocks, and adjusted by leap seconds as needed to keep UTC noon
in synch with astronomical UT1 noon, etc.
Year
The astronomical Julian year is exactly 365.25 days, each day of which has 86,400 SI seconds.
The tropical or solar year is the period of time for the ecliptic longitude of the Sun to increase by 360
degrees...The tropical year is often defined as the time between southern solstices, or between northward
equinoxes [different values result]. Because of the Earth's axial precession, this year is about 20 minutes
shorter than the sidereal year... The mean tropical year, as of January 1, 2000, was 365.241987 days
[each of which has 86,400 SI seconds], or 365 days 5 hours, 48 minutes, 45.1897 seconds... According to
Blackburn and Holford-Strevens (who used Newcomb's value for the tropical year) if the tropical year
remained at its 1900 value of 365.24219878125 days the Gregorian calendar would be 3 days, 17 min, 33
s behind the Sun after 10,000 years... These effects will cause the calendar to be nearly a day behind in
3200.
55

Long time intervals are measured in Ma = 10
6
Julian years, and Ga = Gyr = 10
9
Julian years, each year of
which is 31557600 SI seconds in length. The Gy is the SI abbreviation for Gray, so this abbreviation is
ambiguous for Gigayear. There is no SI abbreviation for year per se. Many persons use yr, such as Myr,
Gyr. These abbreviations may be used for intervals. To express time ago I prefer Mya, Gyae.g., the Big
Bang occurred 13.75 Gya (13.75 Ga ago).

52
http://www.imcce.fr/en/grandpublic/temps/jour_julien.php
53
http://en.wikipedia.org/wiki/Second
54
http://en.wikipedia.org/wiki/Terrestrial_time
55
partly from http://en.wikipedia.org/wiki/Tropical_year
Sky Position, Coordinate System, Motion and Rotation, Ecliptic, and Sidereal Terms

Altitude h
also called elevation, the angle above the local horizon measured along a great circle passing through the
object and the zenith, 0 90. The local altitude of the NCP is the same as local latitude in N hemisphere.
Azimuth A
angle in degrees measured E along the horizon starting at the North point [intersection of the meridian
with the horizon that is closest to the NCP] to the intersection of the great circle used to measure altitude
of the object (0 to 360). Undefined at the Poles.
Celestial Equator CE
projection of the earths equator onto the celestial sphere
Celestial Sphere
including North Celestial Pole NCP, South Celestial Pole SCP, & Celestial Equator CE
Declination
the angle N (+) or S () from the celestial equator for an object in sky
Ecliptic
The apparent path of Sun on the celestial sphere. More precisely, it is the path (the Heliocentric ecliptic) as
seen from the Earth throughout the course of a year of the Suns center. (Or at least it is the path if the
celestial sphere and the Sun could be seen at the same time.) For potentially greater accuracy but more
difficult to compute, the apparent path of the solar systems barycenter (center of gravity, a point which
usually falls within the Sun, see below)
56
may be used to define the ecliptic.
57

Epoch
Positions of celestial objects must be specified in terms of a particular time or epoch, in order to account
for precession, nutation, proper motion, etc. Currently used is epoch J2000.0 (position at noon UT on
January 1, 2000). An earlier system used Besselian years such as B1950.0 as the time of reference.
58

Equatorial Coordinate System
The equatorial coordinate system is widely-used to map celestial objects. It projects the Earth's geographic
poles and equator onto the celestial sphere. The projection of the Earth's equator onto the celestial sphere
is called the celestial equator. Similarly, the projections of the Earth's north and south geographic poles
become the north and south celestial poles, respectively. Coordinates are given as Declination and Right
Ascension RA.
Equinox
An equinox occurs twice a year, when the tilt of the Earth's axis is inclined neither away from nor towards
the Sun, the center of the Sun being in the same plane as the Earth's equator...
59
They are termed Vernal
(March) equinox and autumnal (September) equinoxes... The equinoxes are currently in the constellations
of Pisces and Virgo.
Heliocentric Model of Planetary Motion
Suggested by Nicolaus Copernicus (1473 1543) in his De revolutionibus orbium coelestium, published only
in 1543, near the end of his life, out of fear. Elaborated and expanded by Johannes Kepler (1571 - 1630)
in his Mysterium Cosmographicum, published in 1596, and by Galileo Galilei (1564 1642) in his 1610
Sidereus Nuncius (Starry Messenger) and his 1632 Dialogo sopra i due massimi sistemi del mondo (Dialogue
Concerning the Two Chief World Systems).
Hour Angle HA
Angle W along CE from meridian to hour circle
Hour Circle
The great circle through an object and the celestial poles, therefore perpendicular to the CE.

56
http://en.wikipedia.org/wiki/Ecliptic
57
http://www.astro.sunysb.edu/fwalter/PHY515/glossary.html
58
http://en.wikipedia.org/wiki/Epoch_%28astronomy%29#Besselian_years
59
http://en.wikipedia.org/wiki/Equinox
Horizon
Where the sky meets the ground (in alt-azimuth coordinate system)
Horizon, Horizontal, or Alt/Az Coordinate System
Uses local horizon to measure:
altitude (0 90 above the horizon, also called elevation)
azimuth (0 360 measured from North Point on the horizon toward the E, i.e., to the right)
Meridian
Great circle through local zenith and NCP and SCP. The Prime Meridian is the meridian the runs through
Greenwich, England (0 degrees longitude).
Planetary Positions
Positional terminology expressed relative to the Earth and Sun differ depending on their orbits relative to
the Earth:
Inferior planet: these orbit inside the orbit of the Earth (Mercury, Venus). Relative to Earth, these assume
greatest eastern elongation, greatest western elongation, superior conjunction (in same direction as Sun
but beyond Sun), inferior conjunction (in same direction as Sun but closer than Sun).
Superior planet: these orbit outside that of the Earth (Mars, Jupiter, Saturn, Uranus, Neptune). Relative
to Earth, these assume eastern quadrature, western quadrature, conjunction (in same direction as Sun
but beyond Sun), and opposition (on an extended Earth Sun line in direction opposite to the Sun).
Precession
Arises from tidal forces or torques caused by gravitational forces by Sun and moon on the oblate and tilted
Earth. Earths axial tilt varies between 22.1 and 24.5. Earth precesses over a period of 25770 y, with
current axial tilt angle 23.5 or 23.4. Precession necessitates correction of right ascension and declination
of astronomical bodies and equinoxes etc. relative to the standard reference time when their positions are
tabulated, for instance to epoch J2000.0 (Noon UT in Greenwich England on Jan 1, 2000). (IMA2 p. 13)
Proper Motion and Velocities
Proper Motion is the apparent angular motion transverse to the line of sight between the observer and a
celestial body at distance r. It is expressed as an angular velocity = d/dt = v/r, expressed typically for
stars in arcsec/yr. The linear velocity of the object v is decomposed into the transverse linear velocity v
and the radial linear velocity vr. An angular distance traveled may be expressed in terms of changes in RA
and declination by (IMA2 p. 19):
2
cos
2
+
2

Retrograde Motion
Apparent retrograde motion is the motion of a planetary body in a direction opposite to that of other bodies
within its system as observed from a particular vantage point. Direct motion or prograde motion is motion
in the same direction as other bodies.
60
Retrograde motion was difficult to explain in the Ptolemaic
geocentric universe (named for Claudius Ptolemaeus, c. 90 CE 168 CE, an astronomer who brought the
geocentric model to its near-highest form).
Right Ascension RA or
The angle E along the celestial equator from to the hour circle of a star. It is the angle measured
eastward along the CE from the vernal equinox to its intersection with the objects hour circle (IMA2 p. 12)
It is traditionally measured in hours, minutes, and seconds.
Rotation Direction
When referring to the Earth and the Celestial sphere (see here
61
for animation):
From a Fixed CE Perspective: If the Earth is rotating as viewed looking inward from outside a stationary
Celestial Sphere, the rotation of points on the Earth is to the right (CCW looking down from NCP, CW down
from the SCP) and this rightward direction is toward the Earths east or eastward.
From a Fixed Earth Perspective: If one is on a fixed earth looking out at a rotating Celestial Sphere in the N
hemisphere facing N, the visible CS appears to rotate CCW about the NCP (CW about the SCP in the S
hemisphere), with non-circumpolar objects rising on the right in the east and rotating up and toward the
left then descending in the west. Looking toward the S in the N hemisphere, stars rise on the left (east)
and rotate up and set on the right (west).

60
http://en.wikipedia.org/wiki/Apparent_retrograde_motion
61
http://astro.unl.edu/classaction/animations/coordsmotion/celhorcomp.html
Sidereal Period
The amount of time that it takes an object to make one full orbit, relative to the stars. This is considered to
be an object's true orbital period. Compare synodic period.
Sidereal Time ST
RA + HA. Local sidereal time is the amount of time that has elapsed since the vernal equinox last
traversed the (local) meridian. A mean sidereal day is shorter than a mean solar day, namely
~23h 56m 04.1s = 86164 s
vs.
24*60*60=86400 s
in the mean solar (tropical) day. Thus the mean sidereal day is shorter by 236 s. This is calculated by
86,400s 365.242190402/366.242190402 = 86164 s, because there are 1 more sidereal days than
tropical days in each tropical year.
62
The right ascension of a star is equal to the sidereal time when that
star crosses the meridian... By observation, you could find the sidereal time by locating a prominent star
that is passing directly overhead (crossing the meridian) and then looking up its right ascension on your
star finder or in a star atlas.... For example, the star Capella in the constellation Auriga has the
coordinates RA 5h 13m DEC 46deg. When Capella is directly overhead, the sidereal time is 5:13.
63
See
here
64
for accurate determination of sidereal time.
Solstice
A solstice is an astronomical event that happens twice each year when the Sun's apparent position in the
sky reaches its northernmost or southernmost extremes. The name is derived from the Latin sol (sun) and
sistere (to stand still), because at the solstices, the Sun stands still in declination; that is, the apparent
movement of the Sun's path north or south comes to a stop before reversing direction.
65
They are termed
Summer (or less ambiguously Northern or June) Solstice and Winter (or less ambiguously Southern or
December) solstices.
Their precise times in 2012 in UT are given here:
66

Perihelion Jan 4 at 2:08 Equinoxes Mar 20 at 5:20 Sept 22 at 15:01
Aphelion July 4 at 17:03 Solstices June 20 at 23:21 Dec 21 at 11.23
Synodic period
For planets relative to the Earth, the time interval between successive oppositions or conjunctions, etc.
The synodic period is the temporal interval that it takes for an object to reappear at the same point in
relation to two other objects (linear nodes), e.g., when the Moon relative to the Sun as observed from Earth
returns to the same illumination phase. The synodic period is the time that elapses between two successive
conjunctions with the SunEarth line in the same linear order. The synodic period differs from the sidereal
period due to the Earth's orbiting around the Sun.
67

Vernal Equinox
The point on the CE where the path of the Sun (ecliptic) intersects or crosses from S to N. (The symbol
represents Aries, Latin for ram, not the Greek letter gamma). Also called the First Point of Aries [though
now actually falling in constellation Pisces], aka the ascending node. Also represents the time of this
crossing, typically falling between 3/19 to 3/21. The autumnal equinox falls between Sept. 22 to 24.
Zenith
The point directly over the observers head. In altaz coordinate system, this is alt = 90.

62
http://en.wikipedia.org/wiki/Sidereal_time
63
http://astronomy.uconn.edu/defs/sidereal.html
64
http://aa.usno.navy.mil/faq/docs/GAST.php
65
http://en.wikipedia.org/wiki/Solstice
66
http://aom.giss.nasa.gov/cgi-bin/srevents.cgi
67
http://en.wikipedia.org/wiki/Synodic_period
Celestial Mechanics (Chapter 2)

This chapter has not yet been read in my astronomy courses and only parts of it are summarized here.
Kinematics Defined
Kinematics is the branch of classical mechanics that describes the motion of points, bodies (objects) and
systems of bodies (groups of objects) without consideration of the forces that cause it... The study of
kinematics is often referred to as the geometry of motion... To describe motion, kinematics studies the
trajectories of points, lines and other geometric objects and their differential properties such as velocity and
acceleration.
68

Mathematical Properties of the Ellipse
An ellipse is the set of points satisfying
r + r' = 2a
where a = semimajor axis (1/2 the length 2a of the long axis), and r and r' are the distances of any point on
the ellipse from the two foci F and F', respectively.
The eccentricity e (0 e < 1) is defined as
e (distance between the foci F and F')/2a.
The eccentricity is 0 for a circle. The distance from a focal point to the center of the ellipse (midway between
the two foci) is ae. The semiminor axis is b.

Keplers Original Laws of Planetary Motion

Johannes Kepler (1571 1630) published his first two laws of Planetary Motion in 1609 and his third in
1619.
Keplers First Law
A planet orbits the Sun in an ellipse, with the Sun at one focus. The closest approach of a planet to the Sun
is on the major axis and is termed the perihelion, the furthest being the aphelion.
The distance on an ellipse from the principal focus is given in polar coordinates by
r = a (1 e
2
/ + e cos where e <
and where is the angle measured CCW from the perihelion position (which is taken to be 0).
Keplers Second Law
A line connecting a planet to the Sun sweeps out equal areas in equal time intervals.
Keplers Third Law
P
2
a
3

where P is the orbital period of the planet, and a is the average distance of the planet from the Sun.
For solar planets, if P is measured in sidereal years, and a measured in AU,
P
2
= a
3

68
http://en.wikipedia.org/wiki/Kinematics
Laws of Gravity and Motion

Newtons Law of Universal Gravitation
Isaac Newton (1642 - 1727) published his three laws of motion and the law of universal gravitation in
Philosophi Naturalis Principia Mathematica in 1687:
F = GMm/r
2

where F is the gravitational force between two masses, G the universal gravitational constant (6.673 x 10
-11
N
m
2
kg
-2
), M and m are the masses of the two objects, and r is the distance between them. For extended
masses, F is determined by integrating over the mass distribution(s). For spherically symmetrical mass
distributions, F is directed along the line of symmetry between the two objects and acts as if all the mass of
the spherically symmetrical object were located at the center of the mass. IMA2 proves that, For points
inside a spherically symmetrical distribution of matter, the portion of the mass that is located at radii r < r0
causes the same force at r0 as if all of the mass enclosed within a sphere of radius r0 was concentrated at the
center of the mass distribution. However, the portion of the mass that is located at radii r > r0 exerts no net
gravitational force at the distance r0 from the center.
69

For such a spherically symmetrical body such as a planet, the escape velocity is given by
vesc = (2GM/r)
1/2

where M is the mass of the planet to be escaped from and r is the distance from the center of that planet.

Keplers Laws Updated (by Isaac Newton, etc.)
Define reduced mass as
m1m1/(m1 + m2)
The reduced mass is used in a center-of-mass reference frame in order to reduce the problem to a rotation of
a reduced mass about the total mass located at the origin.
Define as the angle of the reduced mass as measured from the direction to perihelion.
The orbital angular momentum L is conserved in binary orbits.

Kepler First Law, revised: Both objects in a binary orbit move about the center of mass in (separate) ellipses,
with the center of mass occupying one focus of the ellipse. In a center-of-mass reference frame, if L is the
angular momentum (units of kgm
2
s
1
), then the distance r of the reduced mass from the total mass is given
by
r = (L
2
/
2
) / GM + e cos
where M is the total mass of the system, e is the redefined eccentricity, and is as above.
This is the equation of a conic section, an ellipse if the total energy of the system is less than zero (i.e., a
bound system).
In multiple body systems, the barycenter (center of gravity) moves in a more complex manner. For example,
the motion of the barycenter of the Solar System relative to the center of the Sun is depicted in the following
graph, which spans 50 years.
70
Note that the barycenter has not always fallen inside the Sun surface.

69
http://en.wikipedia.org/wiki/Newton%27s_law_of_universal_gravitation
70
http://en.wikipedia.org/wiki/Sun#Atmosphere

Path of Solar System Center of Mass (Barycenter) Relative to the Sun Center, 1945 - 1995

Kepler Second Law, revised:
dA/dt = (L//2
where the right hand side is constant. Presumably (I have not confirmed this) the area is swept out by the
reduced mass and is measured from the center of mass.
Kepler Third Law, revised:
P
2
= [4
2
/ G(m1 + m2)] a

Virial Theorem
The Virial theorem was named in 1870 by Rudolf Clausius (1822 1888), in German fr. Latin, vir = plural
stem of vis: strength, force). For a system that is stably bound by gravitational forces (or by other inverse-
square-law binding forces):
E = K + U = 1/2U
where E is the total energy, <E> is the time-averaged total energy, K is the time-averaged kinetic energy KE,
and U is the time-averaged potential energy PE. As with atoms, the total energy for a gravitationally bound
system is considered to be negative, inasmuch as the components are considered to have zero energy when
infinitely separated, and energy must be radiated away or emitted when the components accelerate inward to
form a bound system. A stably bound system for which the virial theorem applies is said have undergone
virialization and to have become virialized. There are scalar and tensor forms of this useful theorem (and
it has been extended to deal with electric and magnetic fields). It has been used to deduce the presence of
dark matter.

Light Properties (Pre-QM and Early QM); Distance and Magnitudes (Chapter 3)

Astronomical Distance, Luminosity, Magnitude

Stellar Parallax
Astrometry is the branch of astronomy that is concerned with precise determination of 3-dimensional
positions of celestial objects, including their distance.
Stellar parallax is a method of measuring distance to a star (or other celestial object) that is sufficiently
nearby (see image in Chapter 1). The parallax angle p (seconds of arc) is formed by the angle between the
sun, the star, and the earth (note: not the angle formed with the Sun and Earth at opposite orbital locations).
For small angles, distance d is given by
d (in pc) = 1/p where p is in arcseconds
d (in AU) = 206,265/p where 206,265 = arcseconds in a radian, p is in arcseconds
By definition, if p = 1, d = 1 parsec. The nearest star, Proxima Centauri, at 1.3 pc or 4.2 ly has a parallax
angle of only 0.77 arcseconds. The Hipparcos Space Astrometry Mission (HSAM, 19891993) was able to
measure some parallaxes as small as 0.6 mas corresponding to 1600 ly or 500 pc. More accurate missions
are anticipatedGAIA, launching 2013, is expected to attain an accuracy down to 20 as.
71

Luminosity L
is the total energy (J) emitted as electromagnetic radiation or photons (light) per second (thus expressed in
W or J s
-1
). Solar luminosity L varies but is about 3.839 x 10
26
W. (Slightly more energy is emitted if
neutrinos are also included. Particles in the solar wind are not included.) When the term is not otherwise
qualified, luminosity is the same as total bolometric luminosity. (A bolometer is an instrument that measures
radiant energy over a wide band by absorption and measurement of heating.)

Radiant (or radiative) flux F or S
Light from a star emitting uniformly in all directions varies as the inverse square of the distance r (inverse
square law), and is given by F = S = L/4r
2
. (The inverse square law does not apply to a non-spherically-
symmetrical or collimated light beam.) Flux is expressed as power per unit area (W m
-2
) passing through a
plane that is (typically) perpendicular to the direction of radiation travel.
Total Solar irradiance (flux) at 1 A.U. (at the outer Earth atmosphere) varies but averages about 1365 W m
-2

(see later in this document).
Specific Flux density S may also be expressed in Janskys (10
26
watts m
-2
Hz
-1
), named for Karl Guthe
Jansky (1905 1950, one of the pioneers of radio astronomy). Jansky is a non-SI unit approved by the
IAU=International Astronomical Union, and widely used in radio and infrared astronomy. Specific Flux
density incorporates the Hz
-1
unit, and therefore Specific Flux density must be integrated over the finite
receiving band of wavelengths being detected to yield the total flux in this band.
72

71
http://sci.esa.int/science-e/www/object/index.cfm?fobjectid=31197
72
IRA3 p. 9 = Burke BF and Graham-Smith F, An Introduction to Radio Astronomy, 3rd Ed. Cambridge. 2010
Magnitude and Distance Modulus

Apparent magnitude m (m
bol
)
This is the apparent brightness of an object, in which a plus 1 unit increment represents a brightness
reduction by a factor of 100
1/5
= 2.512. The incremental star magnitudes scale derives from the ancient
Greeks, who categorized magnitudes in 6 groups. Smaller (less positive or more negative values) represent
brighter objects. Apparent magnitude here is bolometric (thus measured over all wavelengths of light) and
indicated with mbol, but the term is also used to measure apparent color magnitude (e.g., in blue light, which
is abbreviated mB). The mbol for the Sun is 26.83, and m = +30 for the faintest detectable objects,
representing a ratio of brightness of about 10
23
.
The zero point (zeropoint) of magnitude is approximately the magnitude of Alpha Lyrae or Vega
73
. Specifically,
in the Johnson system, the zero point is chosen such that V = 0.03 for Alpha Lyrae (Vega). But current
definitions for zero point vary and are complicated. For instance, HST states for its Wide-Field Planetary
Camera 2 or WFPC2, The zeropoints in the WFPC2 synthetic system, as defined in Holtzman et al. (1995b),
are determined so that the magnitude of Vega, when observed through the appropriate WFPC2 filter, would be
identical to the magnitude Vega has in the closest equivalent filter in the Johnson-Cousins system. For the
filters in the [WFPC2] photometric filter set, F336W, F439W, F555W, F675W, and F814W, these magnitudes
are 0.02, 0.02, 0.03, 0.039, and 0.035, respectively.
74

For any two magnitudes m1 and m2 (bolometric or color), the ratio of their fluxes is given by
F2/F1 = 100
(m1 m2)/5

Absolute magnitude M
This is the calculated apparent bolometric (or color) magnitude of an object at a distance of 10 pc.

Distance Modulus m - M
The distance to a celestial object is given by
d = 100
(m M + 5)/5
pc
where m and M are apparent and absolute magnitudes. The quantity (m M) therefore serves as a measure
of distance, and is given by
m M = 5 log10(d) - 5 = 5 log10(d/10 pc)
where d is in pc

Comparison with the Sun
The absolute magnitude of a star expressed in terms of luminosity of it and the Sun is
M = MSun 2.5 log10(L/L)
where Mbol,Sun = 4.75 and L= 3.846 x 10
26
W. Here, M is magnitude, not mass.

73
Michael S. Bessell. Annu. Rev. Astron. Astrophys. 2005. 43:293336
74
http://www.stsci.edu/instruments/wfpc2/Wfpc2_dhb/wfpc2_ch52.html
Surface Brightness
For extended objects (such as galaxies, star clusters, or nebulas) that are not point-sources of light, it is
customary to express their surface brightness S (or I) quoted in units of magnitude per square parsec (i.e.,
mag arcsec
-2
). The word surface refers to the fact that the light originates from a spread-out surface on part
of the celestial sphere rather than from a point. For the same emitting object, surface brightness does not
changes with increasing distance (ignoring extinction, etc.)... Since the light is spread out, the average
brightness from any point on this surface is generally much fainter than a star as a point source at
comparable distance.
75

Because the magnitude is logarithmic, calculating surface brightness cannot be done by simple division of
magnitude by area. Instead, for a source with magnitude m extending over an area of A in square
arcseconds, the surface brightness S (quoted in units of magnitude per square arcsecond or mag arcsec
-2
) is
given by
76

+o

Surface brightness S is constant with luminosity distance. For nearby objects, the luminosity distance is
equal to the physical distance of the object. For a nearby object emitting a given amount of light, radiative
flux decreases [by the inverse] square of the distance to the object, but the physical area corresponding to a
given solid angle (e. g., 1 arcsec) increases [also by the square of the distance]..., resulting in the same surface
brightness. The relationship with standard solar magnitude and luminosity units of is given by
e e cseco
+ o
/c

where again the left-hand S of the object is expressed in units of magnitudes per square arcsecond, M is the
absolute magnitude of the Sun in the chosen color band, and the right-hand S of the object is expressed in
units of luminosity of the Sun L [in the same color band?this appears to be a point of confusion] per pc
2
.
A great deal of confusion ensues from the fact that amateur astronomers habitually fail to specify whether
they mean integrated brightness or surface brightness when they say that an object is bright or faint.
Consider, for instance, M33, the Triangulum Galaxy. At magnitude 5.7, it is fifth in integrated brightness of
any galaxy in the sky, after our own Milky Way ... Nonetheless, M33 is referred to as a faint galaxy, because
its light is spread out over a huge areanearly a square degreegiving it one of the lowest surface
brightnesses of any Messier object. On the other hand, the planetary nebula M76 has one of the highest
surface brightnesses of any nebulous Messier object, but it is often called faint because of its low integrated
brightness. (For mathematicians, the term integrated brightness refers to the integral of the surface
brightness over the object's area, which in the case of M76, is tiny.)
77

Light (Photon) Wave vs. Particle Properties

Astronomers tend to use the term light to refer to electromagnetic radiation of any wavelength. Light travels
in vacuum at speed c = 299,792,458 m/s (exactly, by definition). Thomas Young (1773 1829) performed
his double-slit diffraction experiment c. 1803, in which he showed that light constructively and destructively
interferes, thereby establishing its wavelike behavior and deducing a wavelength . Given the finite speed of
light c, the light frequency is given by
c =
He determined the wavelengths of visible light (modern range = 390 to 750 nm, commonly stated as 400 to
700 nm, or 4000 to 7000 ). The range 390 to 750 nm corresponds to frequencies of 400790 THz (1 THz =
10
12
Hz).
James Clerk Maxwell (1831 - 1879) used his Maxwell equations in 1865 to predict electromagnetic transverse
waves propagating at

75
http://www.astro.lsa.umich.edu/undergrad/labs/brightness/index.html
76
http://en.wikipedia.org/wiki/Surface_brightness
77
http://mysite.verizon.net/vze55p46/id18.html
v / 00)
where 0 = permittivity of free space (applying to electric fields) and 0 = permeability of free space (applying to
magnetic fields). These were consistent with light waves.
The Poynting Vector S specified the classical (non-quantum-mechanical) energy (W m
-2
) carried by a light
wave:
S = (1/0) E B
and the vector points in the direction of propagation. The time-averaged value (not root mean square) is
/20) E0B0
although this applies to a specific frequency.
The EM wave carries momentum which can impart a force (i.e., it can exert radiation pressure), given by
Frad /c cos
2
for absorption, or
Frad /c cos
2
for reflection
where A is the area of the surface radiated and is the angle of incidence measured wrt perpendicular to the
plane.
Typically, the radiation pressure arises from Compton scattering of photons by electrons, IMA2 p. 119, and
plays a significant role in the interiors of extremely luminous objects such as early main-sequence stars, red
supergiants, accreting compact stars, and possibly on small dust particles in the ISM. The solar irradiance at
the Earth exerts a pressure of about 4.6 Pa (absorbed)
78
and its effects can be demonstrated with a highly
sensitive Nichols radiometer (1903), but not with the Crookes radiometer (1873).

78
http://en.wikipedia.org/wiki/Radiation_pressure
Spectrum of Electromagnetic Radiation (Photons of Light)
79

Spectrum of Electromagnetic Radiation (Light)
Region
Wavelength
(nm)
Wavelength
(m)
Wavelength
(cm)
Frequency
(Hz)
Energy
(eV)
Radio
(and longer )
> 10
8
> 0.1 > 10 < 3 x 10
9
<1.24x10
5

Microwave
10
8

10
5

0.1
0.0001
10
0.01
3 x 10
9

3 x 10
12

1.24x10
5

0.0124
Infrared
10
5

700
0.0001
7 x 10
7

0.01
7 x 10
5

3 x 10
12

4.3 x 10
14

0.0124
1.77
Visible Light
700
400
7 x 10
7

4 x 10
7

7 x 10
5

4 x 10
5

4.3 x 10
14

7.5 x 10
14

1.77
3.1
Ultraviolet
400
10
4 x 10
7

10
8

4 x 10
5

10
6

7.5 x10
14

3 x 10
16

3.1
124
XRays
(Soft & Hard)
10
0.01
10
8

10
11

10
6

10
9

3 x 10
16

3 x 10
19

124
1.24x10
5

Gamma Rays
(of any origin)
< 0.01 < 10
11
< 10
9
> 3 x 10
19
> 1.24x10
5

Note: opinions differ as to the boundary between soft X-rays and extreme UV, somewhere between 1 & 10 nm.
Many astronomers seem to use 10 nm. Distinction between radio and microwaves also varies.

79
Electromagnetic Spectrum:
http://csep10.phys.utk.edu/astr162/lect/light/spectrum.html (adapted & revised by MCM)
http://en.wikipedia.org/wiki/X-ray
Earths Atmospheric Absorption of Incoming EM Radiation (Light)

Brown regions represent amount of attenuation of the wavelength that has occurred at ground level.
Greatest transparency is seen in visible light, some infrared bands, and at radio wavelengths.
80

Blackbody Radiation and the Planck Distribution Functions
A blackbody is an idealized perfect absorber of incident EM radiation, regardless of frequency or angle of
incidence. A black body in thermal equilibrium (that is, at a constant temperature) emits electromagnetic
radiation called black-body radiation. The radiation is emitted according to Planck's law, meaning that it has
a spectrum that is determined by the temperature alone..., not by the body's shape or composition... A black
body in thermal equilibrium has two notable properties: (1) It is an ideal emitter: it emits as much or more
energy at every frequency than any other body at the same temperature, and (2) It is a diffuse emitter: the
energy is radiated isotropically, independent of direction.. An approximate realization of a black body is a
hole in the wall of a large enclosure... Any light entering the hole is reflected indefinitely or absorbed inside
and is unlikely to re-emerge, making the hole a nearly perfect absorber. The radiation confined in such an
enclosure may or may not be in thermal equilibrium, depending upon the nature of the walls and the other
contents of the enclosure. Real materials emit energy at a fractioncalled the emissivityof black-body
energy levels. By definition, a black body in thermal equilibrium has an emissivity of = 1.0. A source with
lower emissivity independent of frequency often is referred to as a gray body. Construction of black bodies
with emissivity as close to one as possible remains a topic of current interest... In astronomy, the radiation
from stars and planets is sometimes characterized in terms of an effective temperature, the temperature of a
black body that would emit the same total flux of electromagnetic energy.
81

The spectral or specific brightness B(T) (called the Planck function and otherwise unnamed in IMA2, aka
spectral radiance or specific intensity or spectral intensity) for values to be integrated over wavelength and
solid angle differentials is expressed in units of W m
-2
nm
-1
sr
-1
. Note that instead of using pure SI length
units of m
-3
, the nm
-1
term corresponds to the d wavelength differential to be integrated, whereas the m
-2

term corresponds to the area of the emitter. B(T) is reserved for brightness at the emitter surface, whereas
Specific Intensity I , also expressed as W m
-2
nm
-1
sr
-1
, is used for the local value of flux along a ray path,
since it may change with emission and absorption processes (IRA3 p. 9).

80
http://en.wikipedia.org/wiki/Atmosphere_of_Earth
81
http://en.wikipedia.org/wiki/Black_body
The Planck function spectral or specific brightness
may also be expressed as B(T) which is used with
frequency differentials d, or as B(T) which is used
with wave number differentials d (where 1/).
Blackbody radiation is a continuous spectrum with a
peak emission intensity. The hotter a blackbody, the
more the peak is shifted toward the higher energies
(blue, UV, X-ray, etc.) The value of the Planck
function for a particular blackbody and wavelength
(or frequency) is always greater when T is greater,
regardless how much the peak shifts (IMA2 p. 69).
The graph to the right depicts arbitrary intensity
proportional to the spectral radiance (Planck
function) for 3 temperatures, as well as the incorrect
graph derived for 5000 K from classical theory.
82

Wiens displacement law
Derived by Wilhelm Wien (1864 - 1928) in 1893.
83
The wavelength max at maximum or peak blackbody
spectral brightness or intensity B(T) (expressed as W m
-2
nm
-1
sr
-1
) for a given equilibrium blackbody
temperature is given by Wiens displacement law, where b = 2.8978 x 10
-3
m K = 2.8978 x 10
6
nmK
max T = b
max = b / T
Hotter stars appear bluer, cooler stars appear yellower, oranger, or redder. A very hot star having surface
effective T of 42,000 K may has a Wien peak at 69 nm, far into the UV, and cool stars may have peaks in the
infrared. The term displacement refers to the similarity among blackbody curves for differing temperatures,
aside from the differing vertical scale and the displacement laterally of the peak.

Stefan-Boltzmann Equation
This relates Luminosity and Temperature. Deduced by Joef Stefan (18351893) in 1879, and derived from
theoretical considerations, using thermodynamics, by Ludwig Boltzmann (18441906) in 1884.
84
Luminosity
of a blackbody is given by
L = AT
4
where L = luminosity in W, T is temperature in Kelvins, A is area of the blackbody radiating surface, and is
the Stefan-Boltzmann constant, a derived constant = 5.670400 x 10
-8
W m
-2
K
-4
. Clearly, the luminosity is
strongly dependent on and only on the temperature.
For a spherical star or radius R, this gives a relationship between luminosity and effective temperature Te or
Teff:
L =4R
2
Te
4
or
Te = Teff =(L/4R
2
)
1/4

This relationship is used to express an effective temperature, even for objects that are far from having ideal
blackbody spectral distribution. Note that it does not reference color index. The effective temperature is a
property of the surface atmosphere or photosphere, not the interior temperatures.

82
http://en.wikipedia.org/wiki/Planck%27s_law
83
http://en.wikipedia.org/wiki/Wien%27s_displacement_law
84
http://en.wikipedia.org/wiki/Stefan%E2%80%93Boltzmann_law
For the Sun, Luminosity of Sun L = 3.846 x 10
26
W, and Radius R = 6.9599x10
8
m, giving
Teff, = 5777 K

Planck Blackbody Distribution Function (Planck Function)
Max Karl Ernst Ludwig Planck (1858 - 1947) derived his final 1900 formula, which included an empirical
requirement that the allowable energy levels of standing light waves in a blackbody could only be integer
multiples of a minimal energy h or hc/ (i.e., h, 2h, 3h, etc.), where h is a constant, h = 6.6260710
34
J
Hz
-1
. (Eventually it was determined that these energy increments represent individual photons.)
The Planck function expresses the spectral brightness or spectral radiance, given as B(T), B(T), or B(T), thus
as functions of wavelength, frequency, or wave number. B(T) has units of W m
-2
nm
-1
sr
-1
(as discussed
above), and is given by:
T
c

To compute radiant energy per unit time, one must integrate over the range of wavelengths, the area of the
blackbody, and the solid angle into which emission is being calculated.
Integrating this formula over the area of an emitting stars surface gives the monochromatic luminosity L of
the star
The monochromatic flux of a star (joules of starlight energy received per unit area with wavelength between
and + d, and neglecting absorption or scattering) is given by F . (IMA2 p. 75)

Color and Color Index

Stars are viewed through special wide band filters with resulting apparent color magnitudes. The Johnson-
Morgan UBV stellar photometric system (see for example their 1953 paper) uses filters of these color ranges
(IMA2 p. 75):

U = Ultraviolet apparent magnitude mU, with filter centered at 365 nm and effective bandwidth 68 nm
B = Blue apparent magnitude mB, with filter centered at 440 nm and effective bandwidth 98 nm
V = Visual apparent magnitude mV, with filter centered at 550 nm and effective bandwidth 89 nm.
These filters are extended in the Johnson-Cousins-Bessell system to UBVRI, adding:
R = Red apparent magnitude mR, with filter centered at ~640 nm
I = InfraRed apparent magnitude mI, with filter centered at ~798 nm
Additional filters further in the infrared have been used in the 2MASS all sky survey: J (1.25 m), H (1.65
m), and Ks (2.17 m)
85
these aid in the search for cool brown dwarfs. L, M, and N filters have also been
used.
The Sloan Digital Sky Survey SDSS uses u, g, r, i, z (ugriz) filters with effective wavelengths of 360 (UV),
464 (green), 612 (red), 744 (IR), and 890 (IR) nm, respectively. There are a variety of additional systems and
bands sampled. The filters used by HSTs WFPC2 have been previously mentioned.
Absolute color magnitudes MU, MB, and MV may also be calculated if the distance is known.
Commonly used color indices of a star are:
B V = MB - MV
U B = MU - MB

85
http://pegasus.phast.umass.edu/overview/overview.html
These indices are used to quantitate color. The two quantities must be the same type of magnitude, either
apparent magnitudes or absolute magnitudes, yielding equivalent results.
Stars with smaller B V are bluer because lower or more negative magnitudes represent brighter sources.
Stars with larger B V are redder because higher or more positive magnitudes represent dimmer sources.
The color index of a star is a measure solely of the temperature of a model blackbody star and does not
depend on stellar radius or distance (IMA2 p. 77)
The magnitude of Vega has been used as the zeropoint calibration for visual and other color star magnitudes
(see IMA2 p. 76 and earlier discussion of zero point in this document). For each color filter, a constant is
added to the flux measured to give the color magnitude, such that Uvega = 0 For example,
o

(
) +

where SU is a sensitivity of detection fraction for the instrument and CU yields a U of approximately 0 for
Alpha Lyrae (Vega). This does not imply that the actual brightnesses in U, B, and V bands for Vega are
identical. A value Vvega = 0.03 has been chosen for the V band. Other systems of magnitude calibration are
also used which are not tied to a particular star.
The bolometric magnitude is calibrated with
o

(
) +

The Bolometric Correction BC mbol - V. Thus mbol = V + BC. BC is therefore a constant to be added to V to
get mbol. Cbol was originally chosen to keep all values of BC negative, though some supergiant stars have
positive values.
The difference of two magnitudes B1 and B2 comparing two stars is -2.5 log10 (N1/N2), where N1 and N2 are
the total number of [photons] detected from a source above the background sky.
86

A graph of U B versus B V for a population of stars is not a simple straight line due to deviation of these
stars from ideal blackbodies.
Note: There is some confusion when comparing blue-band or other band luminosity of an object to luminosity
of the Sun. It is my understanding according to ZXI that unless otherwise stated, comparisons or ratios are
made comparing:
to

rather than comparing
to

AB Magnitude System
In the AB magnitude system,
87
AB magnitude is defined such that for any bandpass or filter being considered,
the magnitude zero-point corresponds to a flux density of 3631 Jy. (1 Jy = 1 Jansky = 10
-26
W m
-2
Hz
-1

= 10
-23
erg s
-1
cm
-2
Hz
-1
). For monochromatic flux f measured in erg s
-1
cm
-2
Hz
-1

86
Michael S. Bessell. Annu. Rev. Astron. Astrophys. 2005. 43:293336, p. 296
m(AB) = -2.5 log(f) 48.60
The zero point of the AB system uses a flux of 3631 Jy. Conversions with the Johnson system are given in
the cited source. The AB system has been adopted by the Sloan Digital Sky Survey SDSS and the Galaxy
Evolution Explorer (GALEX).

87
AM Magnitude System:
http://en.wikipedia.org/wiki/AB_magnitude
http://www.astro.utoronto.ca/~patton/astro/mags.html#conversions
The Special Theory of Relativity (Chapter 4, omitted)

This chapter and topic has not been included in my astronomy courses to date. However, see my summary
from a Physics 311 course here.
88

Quantum Properties of Light and Matter; Spectroscopy (Chapter 5)

The quantum particle nature of light was postulated by Albert Einstein (1879 - 1955) in his 1905 analysis of
the photoelectric effect. Each metal exhibits a characteristic cutoff frequency c whereby electrons are emitted
only if the frequency of incident light > c . Remarkably, this is independent of brightness. Einstein assumed
photons exist with
E = h = hc/
A 100W bulb emits roughly 2 x 10
20
photons per second. Metals bind electrons, and a minimum binding
energy must be provided to overcome this binding energy.
Spectroscopy

Spectral Lines
In the early 19C, Joseph von Fraunhofer (1787 - 1826) catalogued dark absorption lines in the spectrum of
the Sun, and identified one as arising from sodium.
Gustav Kirchhoff (1824 - 1887) stated his laws of spectroscopy:
1) A hot dense gas or hot solid object produces a continuous spectrum [of light] with no absorption or emission
lines. [This is a continuous blackbody spectrum.]
2) A hot, diffuse (thin) gas produces bright spectral lines (emission lines). [These are produced when the
electron makes a downward transition to a lower orbit with concurrent emission of a photon.]
3) A cool, diffuse (thin) gas in front of a source of a continuous spectrum produces dark spectral lines
(absorption lines) [These are produced when a photon is absorbed causing an upward transition to a higher
orbit.]
Helium absorption lines were first identified in a solar spectrum in 1868, after which the element was found
on Earth.
See below regarding naming of spectral lines.

Doppler Shifts
At non-relativistic slow speeds, approximate Doppler redshift z for the observed wavelength or frequency
relative to the wavelength or frequency at rest at the emitter is given by (per IMA2 p. 112):
z = (obsv - emit) / emit = / emit = obsv / emit - 1 = vr/c, or
z = (femit fobsv ) / fobsv = f / fobsv = femit / fobsv - 1 = vr/c
Here, vr is radial velocity of the emitter with respect to the observer.
Note the denominators, which differ due to the fact that redshift causes wavelengths to increase and
frequencies to decrease
89
. The Doppler shift z is positive if objects are moving apart (redshift to longer

88
http://www.mcgoodwin.net/pages/relativityprimer.html
89
http://ned.ipac.caltech.edu/help/zdef.html
wavelengths) or negative if objects are moving closer (blueshift to shorter wavelengths). See below for
discussion of cosmological redshift, which can no longer ignore relativistic effects.

Light Bending and Spectrographs
Light can be reflected at a surface, in which the angle of incidence equals the angle of reflection.
It can be refracted (by passing from a medium with one index of refraction to one of a different index of
refraction). Such refraction obeys Snell's law: ni sin i = nr sin r, where angle of incidence and of refraction are
measured relative to the normal of the refracting surface.
Light may be dispersed when refracted in a prism or acted on by a diffraction grating. This chromatic
dispersion occurs when the phase velocity and angle of refraction depends on the frequency.
Spectrographs are created with diffraction gratings, which are equivalent to multiple double slits in that they
depend on constructive and destructive interference, and can be either transmissive or reflective. The
smallest wavelength difference that can be resolved is given by /nN where n is the order (1, 2, 3...) of the
spectrum and N is the number of slits. These spectra were originally captured on photographic plates, but
are more commonly now detected by electronic detectors (CCDs).

Zeeman Effect
Splitting of spectral lines due to the Zeeman Effect (Pieter Zeeman 1865 - 1943, described in 1896) is used to
measure stellar and other astronomical magnetic fields. In this phenomenon, the magnetic field resolves
ordinarily unresolved (degenerate) electron configurations having the same energy into slightly different
energies, which are apparent as split spectral lines. The degree of separation is proportional to the magnetic
field except for high field strengths. This allows production of solar magnetograms and even measurement of
weak magnetic fields in the ISM.

Causes of Spectral Line Broadening
Spectral Lines are widened (broadened) by several factors:
90

Natural Broadening by the uncertainty principlegreater broadening E occurs when the excited state is
of shorter duration thus shortening t, as a result of the Uncertainty Principle (Et /2).
Doppler redshift and blueshift effects from rotation, thermal motion, and turbulence
High pressure broadening and collisional broadening, due in part to frequent collisions again increasing
uncertainty E (seen for instance in white dwarf spectra) (IMA2 p. 270)
Zeeman splitting, and
Other inhomogeneities.

Light Interactions and Atomic Models

Compton Effect
Electrons inelastically scatter photons of wavelength i, changing the wavelength in the process according to
the Compton Effect formula:

c
cos

90
IMA2 p. 268 and http://en.wikipedia.org/wiki/Spectral_line

where is the scattering angle of the photon. The electron is scattered into an angle . Clearly, the photon
carries momentum even though it is massless. Arthur Compton discovered this in 1923. The effect is
important because it demonstrates that light cannot be explained purely as a wave phenomenon.

Bohr Hydrogen Atom Model
Niels Bohr 1885 - 1962) developed his semiclassical model of the hydrogen atom (1913), which envisions
electrons in circular orbits and yields a principle quantum number n = 1, 2, 3, ... The allowed energy levels
are -13.6 eV * 1/n
2
. The values of n determine the energy level or electron shell.
Complete ionization of the electron from the ground state requires an input of 13.6 eV into the bound atom.
The various hydrogen emission series are given by the Rydberg formula:
coss (
)
For instance, the Balmer series important in astronomy represents transitions to or from n=2 (which is at
+10.19 eV relative to the ground level). The Lyman series arises from transitions to/from n=1 (which is at 0
eV, the ground level). The Paschen series is to/from n=3 (12.07 eV), Brackett is to/from n=4, and Pfund is
to/from n=5.
A photon emitted from a transition to a lower energy level has energy equal to the difference in energy levels
E = h = hc/f. Similarly, when a photon is absorbed, the energy of the atom is raised by E = h = hc/.
The Bohr model must be refined by fully quantum mechanical considerations with fuzzy orbitals (probability
distributions) and additional quantum numbers.

Series of Spectral Lines
In astronomy, it is common to designate free atoms (not in molecular form) as
Neutral atom H I, He I, C I, etc. (sometimes written HI, HeI, CI, etc., but this is less clear)
Singly ionized ions H II (for H
+
), He II (for He
+
), O III (for O
+
), etc.
Doubly ionized ions He III (for He
++
), O IIII (for O
++
), etc.
(Note that H2 (the diatomic hydrogen molecule) is not the same as H II. In addition, the cryogenic states of
He, termed helium I and helium II (whether as isotopes e
o e
), is not the same as astronomical He I or

He II.)
For unionized atomic hydrogen (termed H I), the commonly encountered spectral lines include those of the
Balmer series, named according to the following pattern
H = 656.281 nm (n = 32 transitions),
H = 486.134 nm (n = 42 transitions),
H = 434.048 nm (n = 52 transitions), and
H = 410.175 nm (n = 62 transitions)
The Lyman lines Ly-, Ly- etc. are in the ultraviolet band. The Paschen and higher n-series lines are in the
infrared. These non-visible light lines are best studied in high-altitude or space-based astronomy, due to
atmospheric absorption.

Quantum Mechanical Considerations

Heisenberg Uncertainty Principle and Quantum Tunneling
The uncertainty (or indeterminacy) principle expresses a fundamental limitation on the simultaneous
knowability of certain paired physical quantities (conjugate variables), and was stated by Werner Heisenberg
(1901 - 1976) in 1927, modified in 1930. The Uncertainty Principle specifies uncertainties E etc. for energy
E, time of determination t, position x, and momentum p (the term uncertainty was not precisely defined):
Et h xp h
Heisenbergs principle was given more precisely by E. H. Kennard in 1927 in terms of standard deviations
and h-bar = = /:
Et / xp /
but he was not describing simultaneous measurements. There remains considerable controversy regarding
interpretation of the Uncertainty Principle.
91

The uncertainties in position allow particles or photons to cross (tunnel through) barriers in which
classically they cannot exist. Tunneling
92
through a potential barrier is more likely when the barrier is no
more than a few wavelengths wide.

Wave-Particle Duality
The wavelike aspects of particles (wave-particle duality) were proposed by Louis de Broglie ( 1892 1987) in
1927. As with light photons, particle frequency is given by
= E/h
and particle wavelength is given by
= h/p
where p is the momentum. This hypothesis was subsequently confirmed in double slit experiments. For
electrons,
= h/mev
A sample calculation for a free electron moving at 0.01c shows the wavelength to be 250 pm, which is similar
to the diameter of an unbound hydrogen atom (100 pm).

Schrdinger Equation and the Quantum Mechanical Atom
The waves de Broglie postulated are considered to be waves of particle positional probability, which are given
by ||
2
. The function is the Schrdinger equation, developed in 1925 by Erwin Schrdinger (1887 - 1961).
Destructive interference of two particles (locations where interfering particles are not found) is seen where
|1 + 2|
2
= 0.
The Quantum Mechanics (QM) model of the atom introduces three additional quantum numbers. The
angular momentum vector for the atom L is determined by l (the azimuthal or angular momentum quantum
number), which can take on values of l = 0, 1, 2, ..., n-1. The quantum number l corresponds to the
subshell orbitals, which have been named by the l values: s (l=0), p (l=1), d (l=2), f (l=3), g (l=4), h (l=5), etc.)
The magnitude of L can take on only a set quantized levels determined by l (IMA2 p. 133):
+
The quantum number ml (magnetic quantum number) determines possible values of the angular momentum
z-component Lz. It specifies the specific orbital or cloud within the subshell, and can only take on values
limited by l. (The following formula contains l, not the digit 1):

91
http://en.wikipedia.org/wiki/Uncertainty_Principle
92
http://en.wikipedia.org/wiki/Quantum_tunnelling
The quantum numbers l and ml have no effect on the energy of a free isolated hydrogen atom.
However, if a magnetic field is present giving a preferred direction, the normally degenerate configurations
(which all have the same energy and therefore produce the same spectral lines) acquire slightly differing
energy levels due to differing values of ml. The result is that split spectral lines can arise from the differing
ml values. Typically, three split lines are produced, not all of which are evident from a particular viewing
angle (due to polarization effects). Thus, the line with unshifted frequency 0 corresponding to ml = 0 may not
be visible in a sunspot when the magnetic field is directed parallel to the line of sight, and only shifted lines
are seen. As previously mentioned, this splitting is termed the Zeeman Effect. Even the ISM has a faint
magnetic field, as low as 2 x 10
-10
T, but this can be detected by measuring the Zeeman line shifting of
hydrogen emissions.
A fourth quantum number ms determines the orientation of the spin angular momentum of the electron. The
overall magnitude of the spin angular momentum vector S is fixed for a particular particle, and S is (
) for
the electron. The allowed values for the magnitude of the electron spin z-component Sz (where z direction is
determined by the direction of the prevailing magnetic field) are Sz = ms = (1/2). Thus the quantum
number ms may take on values of 1/2, depending on whether Sz is spin up or spin down. If the
orientation of the electron spin angular momentum vector Sz is up or parallel to the z-component of the
protons angular momentum vector Lz, it is at slightly higher energy (less unbound), compared to the
antiparallel orientation which is slightly lower energy.
This leads to an additional potential cause of spectral line splitting, due to this anomalous Zeeman Effect.
Such hyperfine splitting of the ground state is the source of the 1420.406 MHz 21 cm H line.
Note that for the H ground state, n=1, l=0, ml = 0, and mS= 1/2. This state may be represented
93
by
symbols H(1s
1
) or the Russell-Saunders atomic term symbol
2
S1/2.

Pauli Exclusion Principle
Wolfgang Pauli (1900 - 1958) deduced in 1924-5 that no two electrons can occupy exactly the same quantum
state, defined by the four quantum numbers (see above). This principle was derived by Paul Dirac (1902 -
1984) using relativistic wave equations. He showed further that this principle applies to all fermions. These
are protons, neutrons, quarks, and all leptons including electrons, muons, taus, and neutrinos, etc.
Fermions all have spins which are half-integer multiples of , namely /2, 3/2, 5/2, etc.
94
(In contrast,
Bosons have spins that are integer multiples of , namely 0, , 2, 3, etc., and they do not obey the Pauli
Exclusion Principle.) See elsewhere in this document regarding degeneracy pressure.

Forbidden Transitions
Larger atoms have many possible energy levels and complex spectra result. Complex selection rules lead to
forbidden vs. allowed transitions. Forbidden transitions may still occur but with much lower probability. For
instance, the 21-cm line for Hydrogen is a forbidden transition, having a probability
95
of only 2.910
15
s
1
.
This 1420 MHz radiation arises from the transition between the two levels, F=1 and F=0, of the hydrogen
ground state, which are slightly split by the interaction between the electron spin and the nuclear (proton)
spin. The parallel (aligned) state of spin is slightly higher in energy (thus less loosely bound) than the
antiparallel (anti-aligned) state.
In astronomy, forbidden line transitions are notated by placing square brackets around the atomic or
molecular species in question, e.g., [O III] or [S II].

93
Atomic Symbols and Russell-Saunders Term Symbol:
http://www.chem.ufl.edu/~itl/4412/lectures/ATermSym.html
http://en.wikipedia.org/wiki/Term_symbol
94
http://en.wikipedia.org/wiki/Fermion
95
http://en.wikipedia.org/wiki/Hydrogen_line
Telescopes (Chapter 6)

This is a limited survey of Chapter 6.
These topics are not summarized: Refraction, Reflection, Snells Law of Refraction, Focal Plane, Spherical
Aberration, Chromatic Aberration (with refractors), Coma, Astigmatism, Field Curvature, Distortion,
Magnification, Refracting and Reflecting Telescope Designs, Telescope Mounts, Large-Aperture Telescopes,
Adaptive Optics, CCDs, Radio Telescopes, IR, UV, X-ray and Gamma ray telescopes, and All-Sky Surveys and
Virtual Observatories.
Viewing Conditions and Site Selection
Astronomical viewing conditions (seeing conditions) improve with reduction of interposed atmospheric
turbulence (which causes twinkling, water vapor, other absorbing gases, dust, etc. The HST Hubble Space
Telescope addresses these needs, as will the James Webb Space Telescope JWST.
Ground based observatories attempt to locate at high altitude in dry locations: on Mauna Kea, these include
Keck I and II, Gemini Northern, the Canada-France-Hawaii CFHT, Subaru, Univ. Hawaii, NASA Infrared IRTF,
United Kingdom Infrared UKIRT, Caltech Submillimeter Obs. CSO, James Clerk Maxwell JCMT, the
Submillimeter Array (SMA) and the Very Long Baseline ArrayVLBA. Other preferred locations for current and
future instruments include the Atacama Desert, Chile and other Andean locales.

Focal Point and Focal Length f
Focal length is the distance between the center of a refracting objective lens or curved reflective mirror surface
and its focus (focal point). The focal point is the point at which incoming light rays parallel to the optical axis
are focused by the lens. Focal length is a measure of how strongly the lens converges or diverges light.

Plate Scale
For astronomical objects, the plate scale d/dy = 1/f is defined for each telescope and relates the angular
separation d of the celestial sphere of an observed object to its linear separation dy at the focal plane of the
detector. is in radians, f is the focal length, y and f are in identical units. This is the IMA2 definition p. 144.
Plate scale may also be defined as the number of arcseconds that each pixel of the CCD detector spans
(subtends).
96
In this case:
plate scale (arcsec/pixel) = s / d
(where s is individual pixel size and d is overall CCD detector size.
Then
plate scale (arcsec/pixel) = s/d = s/(f/206265) = 206265 s/f
where pixel size s and focal length f are in the same length units.

Magnification
For traditional refracting telescopes, when the objective and eyepiece lenses are separated by the sum of their
focal lengths, angular magnification of the virtual image is the ratio fobj (the telescope objective focal length) to
feye (the eyepiece lens focal length, IMA2 p. 154, 175). This formula also applies to reflectors, but
magnification is less useful especially for CCD detector systems, and plate scale is presumably a better index.

Resolution
Is based on an analysis of Airy disk interference for point sources. It specifies that two point sources are
unresolved if the central maximum of one points Airy pattern falls inside the location of the first minimum

96
http://www.handsonuniverse.org/for_teachers/course/teacher_course/size2/size23.html
(circle of lowest intensity) of the other Airy pattern. This requirement leads to the Rayleigh Criterion for
angular resolution:

where is the minimum resolvable angle in radians of the object as viewed under ideal seeing conditions
without a telescope, is the wavelength, and D is the telescope aperture diameter in the same units. Clearly,
resolution improves with shorter wavelengths and wider apertures, all other factors being unchanged.

Focal Length and Focal Ratio (f-ratio)
The focal ratio F is given by f/D (where D is the aperture diameter). It also is called the f-ratio, f-stop, the f/
number (written preceded by f/, such as f/2.8), or the relative aperture. It is a unitless quantity as it is a
ratio of identical length units.

Image Brightness
The image intensity of light (in W m
-2
nm
-1
sr
-1
) from astronomical sources typically falls off only from
attenuation and scattering, and otherwise the image intensity is equal to the object intensity (a somewhat
surprising result of the geometric relationships and definitions, IMA2 p. 153).
However, telescope light gathering effectiveness is expressed by illumination J, the amount of light energy (or
number of photons) focused per unit time per unit area of the resolved image at the detector. The
illumination J is proportional to 1/F
2
or (D/f)
2
, that is the inverse square of the f-ratio. If illumination is
reduced in half, twice the time is required to gather the same number of photons. Comparing two telescope
systems having f-ratios of 2 and 4, the one with F=2 has 4 times the illumination at the detector as the one
with F=4.

Effect of Modification of Telescope Design Parameters
Increasing telescope aperture D (keeping focal length f unchanged) increases illumination J and improves
resolution (decreasing min).
Increasing focal length f (keeping D unchanged) reduces illumination J but increases image size on the
detector (spans more pixels).
Increasing telescope aperture D (while keeping the focal ratio F unchanged, therefore increasing f
proportionately) keeps illumination J constant, but improves resolution (lower min).
Increasing focal length increases telescope support mass and complexity.
Increasing aperture D increases mirror weight and potential for distortion, necessitates greater mirror
support.

Binary Systems and Stellar Parameters (Chapter 7)

Binary star systems are widespreadmore than 1/2 of all stars are in multiple systems, in which two or more
stars are gravitationally bound and orbit about a common center of mass COM or barycenter). Some binary
systems are useful for determining star masses and sizes. When differing in luminosity, etc., the more
luminous is often termed the primary star, the other a companion star. Binary systems may be classified as
follows:
[Optical Double: Not a binary system, simply two stars juxtaposed near a line of sight, but not
gravitationally bound.]
Visual Binary: Two stars optically resolvable due to adequate angular separation.
Astrometric binary: A true binary where the primary is much brighter, and the unseen companion is
detected only by oscillatory motion induced in the position of the primary star about the COM. This method is
also used to detect extrasolar planets.
Eclipsing binary: a binary with orbital plane approximately parallel to the line of sight, and for which the
transit of one star in front of the other is detected by means of characteristic changes in the light curve. The
effective temperatures and radii may often be inferred.
Spectrum binary: a binary detectable by means of two superimposed independent discernible spectra.
Separate blackbody-like spectra are usually detectable only when the cooler star is much larger in size (such
as R Aquarii).
Spectroscopic binary: If the stars have comparable size and luminosity, Doppler effects may cause
detectable changes in the spectral lines of both stars (a double-line spectroscopic binary). When one stars
lines are redshifted the others will appear blueshifted due to opposing velocities in the orbital plane about the
COM (see demo
97
). If one star is much more luminous, the lines of the companion may be obscured but its
presence may be detected by periodic fluctuation in the primarys lines.
Some binaries may satisfy more than one type named above, e.g., an eclipsing spectroscopic binary.
Binary systems may allow star mass determination in certain circumstances: visual binaries close enough to
allow parallax distance determination, visual binaries with radial velocities available for full orbits; and
eclipsing double-line spectroscopic binaries.
Many of the same kinds of techniques described for binary systems are now also used in the search for
extrasolar planets.

Mass Determination Using Visual Binaries

To compute mass ratios: Two stars resolvable by adequate angular separation allow determination of the
orientation of the orbits and the COM. The ratio of the masses is given from the semimajor axes if they are
known:

where m1 and m2 are the stellar masses and 1 and 2 are the respective angular size of the semimajor axes
of the two elliptical orbits about the COM. Thus the mass ratio can be computed even if the distance is
unknown (IMA2 p. 184). Within limits, the orbital inclination plays no role in computing this ratio.
To compute individual masses: If distance d is known (by whatever means), the linear separation of the
stars can be computed. Here the orbital inclination i, measured wrt a plane perpendicular to the line of sight,
must be estimated. A relatively simple formula applies when the orbital plane has an inclination wrt the
plane of the sky but there is no tilt about the major axis or rotation of the ellipse about the line of sight (IMA2
p. 185). Keplers third law as modified may then be used to provide the sum of the masses:
cos
)
cos
cos
)

97
http://www.astro.cornell.edu/academics/courses/astro101/java/binary/binary.htm
provided
, the semimajor axis of the orbit of the reduced mass, the inclination i is known, and
the special conditions on the tilt as mentioned apply. The values for the ratio and sum of the masses may
then be solved for the individual masses m1 and m2. The possibility of additional types of angulation of the
orbital plane makes the computation more complex.
Solutions for mass and other binary stellar parameters are often made with computerized data fitting
techniques, comparing a synthetically generated light curve based on postulated exact parameters to the
actual light curve and adjusting the synthetic parameters until the best fit is obtained.
Even if the distance is not known, it may still be possible to determine individual masses using Keplers 3rd
law if detailed radial velocity data are available.

Eclipsing Spectroscopic Binaries

Double-line eclipsing spectroscopic binary systems are especially useful for computing star parameters such
as their masses, radii, ratio of their fluxes, and ratio of their effective temperatures (IMA2 p. 186).

Mass Determination
The radial velocities of the individual stars, which determine the spectral line shifts, are influenced by the
inclination, the ellipticity, the proximity, the rotation about the line of sight, etc.
See this demo
98
of radial velocities to better understand the effects of these factors. (In the demo, a is the
orbital separation in solar radii determining period, e is the eccentricity, i is the inclination angle as seen from
Earth, w is the node anglethe angle of the major axis as measured in the orbital plane, probably the same
as orientation of the periastron).
Close binaries tend to circularize their orbits over short relative time scales due to tidal interactions. Stars in
circular orbits have constant speed along their orbits given by v1=2a1/P, where P is period and a1 is the
semi-major axis of star 1. The ratio of star masses is given as
/s
v
/s

where v1r is the observed radial velocity for star 1. Thus the ratio of masses is independent of the inclination.
To obtain the sum of the masses, we must know the inclination i. IMA2 p. 188 derives:

where the right hand side is the mass function requiring only the easily determined period and radial velocity
of the primary star. Using this and the previous equation, masses can be determined if radial velocities of
both and the inclination are obtainable. Even if only m1 can be obtained, the mass function sets a lower limit
for m2. As inclinations are often not known, these formulas are sometimes used for statistical studies giving
mass estimates for a class of stars defined by luminosities and effective temperatures, etc.
In general, there is a mass-luminosity relationship for Main Sequence stars that has been derived using
masses determined with visual and double line spectroscopic binaries. The empiric relationship is:

98
http://www.astro.cornell.edu/academics/courses/astro101/java/binary/binary.htm
where the exponent is found to be 3 < a < 4, quoted as 3.5 for 2M < M < 20M. Data showing this
relationship is graphed below:
99

MassLuminosity relationship for Main Sequence Stars

OB = Optical Binaries (data of Popper 1980, image per footnote)

Inclination
Eclipsing binaries in most cases must have i (inclination) close to 90 relative to the plane of the sky. The
specific pattern of light curvesspecifically whether the minima are constant (aside from limb darkening
effects) or curvedallows determination whether the geometry is partially or fully eclipsing and thereby
improves the estimate of i.

Stellar Radii
The star radii may be determined from the amount of time between first contact ta and the point of minimum
light tb as well as the star velocities. (IMA2 p. 191):
rs = (v/2)(tb-ta) for the smaller star, etc., where v = vs + vl, the relative velocities of the two stars

99
image from http://faculty.physics.tamu.edu/belyanin/astr314/lecture8.pdf (A. Belyanin), similar to IMA2
p. 189, adapted from data from Popper DMA, Rev. Astr. Astrophys. 18, 115 (1980)
Ratio of Stellar Effective Temperatures T
eff

The ratio of the effective temperatures may be obtained from these light curves by assuming the stars are
blackbody radiators (IMA2 p. 192). The transit of a less luminous star in front of a more luminous star
creates a deeper dip in the light curve than the opposite transit. In the textbook example, the ratio of
effective temperatures may be derived from:
(
T

where B0 is the brightness when both stars are fully seen, Bp is brightness of the primary minimum (when the
smaller hotter star is eclipsed), and Bs is the brightness of the secondary minimum. Ts is the effective
temperature of the smaller star.

The Classification of Stellar Spectra (Chapter 8)

Harvard Classification

Stellar classification by spectra began with Harvards Edward Pickering (1846 - 1919) and Williamina P.
Fleming in the 1890s using capital letters, as well as by Antonia Maury. Astronomer Annie Jump Cannon
(1863 - 1941) gave us the final sequence of the first 7 letters O B A F G K M, and their subdivision into A0 -
A9 etc., based on the strength of the hydrogen Balmer lines. Her observations on 20,000 stars were compiled
into the Henry Draper Catalogue, from which stellar HD numbers derive. This taxonomy is termed the
Harvard Classification,
100
and is based (as became clear only later) on a sequence of decreasing stellar surface
temperatures. The currently expanded sequence of letters is O B A F G K M R N S L T etc., for which there
are some amusing mnemonics.
101
(The letter P in the Draper catalog referred to Planetary Nebula, and Q were
other spectraletters omitted currently.) O stars are the hottest (and therefore newest), with surface
temperatures of 33,000 K. L and T are very cool stars and brown dwarfs. Our Sun is a G2V star (V is the
luminosity class, see below). The stars near the beginning (O B A) are still termed early-type and the stars
near the end (K M) are late-type, terminology reflecting now-discredited models of stellar evolution (IMA2 p.
220).

The Harvard Spectral Classification
102

Data shown for M
, R
, and L
apply to main sequence [luminosity class V] stars only

Class Surface
Temp K
Conven-
tional
color
Apparent
color
Mass
M

Radius
R

Luminosity
(Bolometric)
L

Hydrogen
Lines
% of Main
Sequence
Stars
O 33,000 blue blue 16 6.6 30,000 Weak ~0.00003%
Hottest blue-white stars, few lines, Strong He II absorption (sometimes emission), He I lines becoming
stronger.
B 10,000
33,000
blue to
blue white
blue
white
2.116 1.86.6 2530,000 Medium 0.13%
Hot blue-white, He I absorption lines strongest at B2, H I n=2 (Balmer) abs. lines becoming stronger.
A 7,500
10,000
white white to
blue
white
1.4
2.1
1.41.8 525 Strong 0.6%
White, Balmer abs. lines strongest at A0, becoming weaker later. Ca II abs. lines becoming stronger.
F 6,000
7,500
yellowish
white
white 1.04
1.4
1.151.4 1.55 Medium 3%
Yellow-white. Ca II lines continue to strengthen as Balmer lines continue to weaken. Neutral metal abs.
.lines (Fe I, Cr I)

100
http://en.wikipedia.org/wiki/Stellar_classification#Harvard_spectral_classification
101
OBAFGKMRNSLT mnemonic:
Oh be a fine girl, kiss me
Oh Brother, Astronomy Finally Gruesomely Killed Me Right Now (Slump)
Oven Baked Ants, Fried Gently, Kept Moist, Retain Natural Succulence (Largely True)
102
Stellar Classification:
http://en.wikipedia.org/wiki/Stellar_classification
IMA2 p. 205
faculty.physics.tamu.edu/belyanin/astr314/lecture8.pdf
G 5,200
6,000
yellow yellowish
white
0.8
1.04
0.96
1.15
0.61.5 Weak 7.6%
Yellow. Sun is G2. Ca II lines continue becoming stronger. Fe I, other neutral metal lines becoming stronger.
K 3,700
5,200
orange yellow
orange
0.45
0.8
0.70.96 0.080.6 Very weak 12.1%
Cool orange. Ca II H and K lines strongest at K0, becoming weaker later. Spectra dominated by metal abs.
lines.
M 3,700 red orange
red
0.45 0.7 0.08 Very weak 76.45%
L Very cool, dark red (brown dwarfs). Brighter in Infrared than visible. Strong molecular absorption
bands of metal hydrides (CrH, FeH), water, CO, and alkali metals (Na, K, Rb, Cs). TiO and VO
weakening
T Coolest stars or brown dwarfs. Strong methane (CH4), weakening CO bands.

The letters R N and S were for peculiar stars (with high amounts of C, lanthanum oxide, or zirconium oxide,
etc.) and are no longer used. There are several published catalogs of stellar spectra.
103

Other Libraries and Models of Stellar Spectra
In addition to the textbook (IMA2 p. 205), may sources graph, tabulate, and compare the physical
characteristics of stars emitting the various spectra
104
I am yet to make a careful study of this huge body of
information. ZXI commends the Robert L. Kurucz models of Stellar Energy Distributions (SEDs) [aka Spectral
Energy Distributions]
105
as a means of understanding how these spectra arise. He states that the Kurucz
models make use primarily of the effective temperature, surface gravity, and metallicity.

Statistical Mechanical Considerations

A fuller understanding of the source of these spectra required analysis not only of electron orbital
probabilities and also the probability of ionization given by statistical mechanics.
The Maxwell-Boltzmann velocity distribution function (IMA2 p. 206 and here
106
), the most probably speed of
an atom vmp=(2kT/m)
1/2
, and the root mean square speed of an atom vrms=(2kT/m)
1/3
are illustrated as follows
for hot hydrogen atoms (graph does not yet consider the likelihood of ionization):

103
Jacoby, G. H., Hunter, D. A., & Christian, C. A., A library of stellar spectra, Astrophysical Journal
Supplement Series (ISSN 0067-0049), vol. 56, Oct. 1984, p. 257-281, available online in PDF form.
www.annualreviews.org/doi/pdf/.../annurev.astro.42.053102.134017 for L and T stars & brown dwarfs
104
105
Sample Kurucz papers and models of Stellar Energy Distributions (SEDs):
http://www.mso.anu.edu.au/observing/2.3m/CASPIR/docs/manual/node228.html
Kurucz, R. L., Model atmospheres for G, F, A, B, and O stars, Astrophysical Journal Supplement Series, vol.
40, May 1979, p. 1-340.
R. Buser and R.L.Kurucz. A library of theoretical stellar flux spectra. I. Synthetic UBVRI photometry and
the metallicity scale for F- for K-type stars A&A 264, 557-591 (1992) [340 pages long!]
106
http://en.wikipedia.org/wiki/Maxwell%E2%80%93Boltzmann_distribution

MaxwellBoltzmann Probability Density Function for Hydrogen Atoms at T = 10,000K
The fraction of hydrogen atoms in the gas having velocities between 2x10
4
m s
-1

and 2.5x10
4
m s
-1
is the shaded area under the curve between those two velocities
divided by the entire area under the curve.
107

The Boltzmann Equation
is given as follows:
/

where P(Sb) is the probability that a system is in state Sb, Eb is the energy of this state (based on the quantum
numbers characterizing this state), T is the temperature in K, and k is the Boltzmann constant. This
equation shows that higher energy orbitals are statistically less likely to be occupied than lower energy
orbitals, all other factors being equal.
The important and recurring quantity
e
/

is called the Boltzmann factor. At room temperature (300 K), kT = 1/40 eV.
States Sb, and Sa, with equal energies (Eb = Ea) but which are not identical (having identical quantum
numbers) are termed degenerate. The presence of multiple degenerate states when present must be
accounted for in the expression of the Boltzmann equation.
The final Boltzmann Equation applied to stars (for which probabilities are proportional to numbers of atoms),
and including g factors allowing for degeneracy, and again not accounting for ionization, is given as follows
(IMA2 p. 212):
/

where Nb is the relative number of atoms in state Sb, Eb is the energy of this state (based on the quantum
numbers characterizing this state), T is the temperature in K, and k is the Boltzmann constant.

107
http://faculty.physics.tamu.edu/belyanin/astr314/lecture8.pdf after IMA2 p. 211.
If one attempts to use this equation to find at what temperature there are equal numbers of unionized H
atoms in the ground state (n=1) and the next higher state (n=2), the answer computed is 85,400K. This is
incorrect, because ionization has not been considered (see to follow).

Saha Equation
Developed in 1920 by astrophysicist Meghnad (or Megh nad) Saha (1894 - 1956), this gives the relative
number of atoms in different states of ionization. Ionization energies needed to go from state i to next higher
level of ionization i + 1 are represented by i. The ratio of the number of atoms in ionization stage (i + 1) to
the number in state (i) is given by:
)
/
e
/

where me is the electron mass, Pe is the pressure of free electrons (which can often only be estimated), and the
Zi terms are partition functions incorporating degeneracy factors gj:

e
(
)/

For stars, electron pressure in the atmospheres ranges from 0.1 to 100 N m
-2
.

Computing the Strength of Balmer Lines versus Calcium Lines, etc.
In practice, the Boltzmann and Saha equations are combined to compute stellar spectra, for instance by
computing the relative numbers of n=2 hydrogen atoms available to create Balmer lines. For example, Saha
shows us that for H, 5% of atoms are ionized at 8300K, 50% are ionized at 9600K, and 95% are ionized at
11,300 K. This 3000K range (from 5% to 95% ionization) is termed the partial ionization zone, and is
conveniently approximated for H as 10,000K. The Balmer lines (H, H, etc.) are seen maximally at 9250 K
because at temperatures higher than 10,000K, H is nearly fully ionized so cannot produce Balmer lines. The
population of H in the state n=2 peaks at 9,900 K (assuming Pe = 20 N m
-2
), and at higher temperatures there
is a rapid falloff of n=2 H I atoms to provide Balmer lines. When other atoms are present, such as He, there
are additional electrons to recombine with ionized H, so a higher temperature is required to achieve the same
degree of H ionization.
After computing with Saha Equation the fraction on atoms which are in a particular state of ionization, one
uses the Boltzmann Equation to obtain the number of atoms in a particular excitation state available to
create spectral lines.
For the Sun with Te = 5777 K, only a tiny fraction of H is in the n=2 H I state, and Balmer lines are
inconspicuous, whereas Ca II H and K lines are strong (IMA2 p. 219).

Compositions of Stellar Atmospheres
Cecilia Payne (1900-1979) was the first to calculate the relative compositions of stellar atmospheres in her
1925 thesis.
Here is a table of the 10 most common elements in the Sun (apparently based on solar spectra)
108
:

108
http://imagine.gsfc.nasa.gov/docs/ask_astro/answers/961112a.html See also IMA2 p. 277
Element Solar Abundance
(% of total
number of atoms)
Solar Abundance
(% of total
mass)
Hydrogen 91.2 71.0
Helium 8.7 27.1
Oxygen 0.078 0.97
Carbon 0.043 0.40
Nitrogen 0.0088 0.096
Silicon 0.0045 0.099
Magnesium 0.0038 0.076
Neon 0.0035 0.058
Iron 0.030 0.014
Sulfur 0.015 0.040

The nine most abundant cosmic nuclear species in descending order are: hydrogen-1, helium-4, oxygen-16,
carbon-12, neon-20, nitrogen-14, magnesium-24, silicon-28, and iron-56 (IMA2 p. 315)

Here is a chart showing the number abundances of the elements in the solar system (not just the Suns
photosphere):
109
The textbook (IMA2 p. 541) shows a similarly plotted graph for the relative abundance of the
Suns photosphere. H and He are of course much more abundant there.

The following graph indicates the relative spectral line strengths of various elements and the TiO molecule in
stellar atmospheres as a function of spectral type or temperature
110
:

109
http://en.wikipedia.org/wiki/Abundance_of_the_chemical_elements
110
http://faculty.physics.tamu.edu/belyanin/astr314/lecture8.pdf adapting IMA2 p. 220

The Hertzsprung-Russell (H-R or HD) Diagram

The data was compiled in tabular form c. 1910 by Ejnar Hertzsprung (1873 - 1967) and Henry Norris Russell
(1877 - 1957), initially as a means of studying stellar evolution. Russell published the information in
graphical form in 1913. An excellent simulation of stellar evolution on a H-R diagram grid is given here
111
. A
modern version follows:

111
http://www.astro.ubc.ca/~scharein/a311/Sim/hr/HRdiagram.html requires JAVA

A modern HertzsprungRussell diagram
112

with 22,000 stars plotted from the Hipparcos catalog
and 1,000 from the Gliese catalog of nearby stars.
The Sun, a G2V star, is located at absolute magnitude 4.8, luminosity 1, and B-V=+0.66
The ordinary hydrogen-burning dwarf stars like the Sun are found in a band running from top-left to bottom-
right called the Main Sequence. Giant stars form their own clump on the upper-right side of the diagram.
Above them lie the much rarer bright giants and supergiants. At the lower-left is the band of white dwarfs
these are the dead cores of old stars which have no internal energy source and over billions of years slowly
cool down towards the bottom-right of the diagram.

112
HertzsprungRussell diagram:
image and most text that follows it from http://www.atlasoftheuniverse.com/hr.html (Richard Powell)
a few comments also from http://en.wikipedia.org/wiki/Hertzsprung%E2%80%93Russell_diagram
Stellar Spectral Energy Distributions (SEDs) are a function of mass, chemical composition, and age. Kurucz
models (1979) assume main sequence star SEDs are determined by effective temperature, log(g), and
metallicity (expressed as [Fe/H] (ZXI notes).
The original [H-R] diagram displayed the spectral type of stars on the horizontal axis and the absolute
magnitude on the vertical axis. The first quantity (i.e. spectral type) is difficult to plot as it is not a numerical
quantity, and in modern versions of the chart it is replaced by the B-V color index of the stars.
113

After stars have consumed much of their H, they turn off the main sequence (a period called the Post-Main
Sequence).
Morgan-Keenan Luminosity Classes and the M-K Classification System
In addition to the Harvard Classification of star spectral classes, stars are also classified into five main
luminosity classes. Stars having the same spectral type (thus same effective temperature), were noted by
Antonia Maury and Hertzsprung to vary in line width. This observation culminated in the 1943 publication of
the Atlas of Stellar Spectra (MKK Atlas) by William W. Morgan (1906 - 1994), Phillip C. Keenan (1908 - 2000)
and Edith Kellman (1911 - 2007), all of Yerkes Observatory. Their MKK atlas and classification adds a
luminosity class (designated by Roman numerals I to V) to the Harvard classification spectral type. For stars
of the same spectral type, narrower lines (lower Roman numeral M-K luminosity classes) are seen with stars
of higher luminosity. (This is because they have thinner atmospheres and their atoms experience fewer
atomic collisions, causing less collisional broadening of lines.)
These are the five luminosity classes (quoted from here
114
):
I Supergiants
Very massive and luminous stars near the end of their lives. They are subclassified as Ia or Ib, with Ia
representing the brightest of these stars. These stars are very rare - 1 in a million stars is a
supergiant. The nearest supergiant star is Canopus (F0Ib) 310 light years away. Some other examples
are Betelgeuse (M2Ib), Antares (M1Ib) and Rigel (B8Ia).
II Bright Giants
Stars which have a luminosity between the giant and supergiant stars. Some examples are Sargas
(F1II) and Alphard (K3II).
III Normal Giants
These are mainly low-mass stars at the end of their lives that have swelled to become a giant star. This
category also includes some high mass stars evolving on their way to supergiant status. Some
examples are Arcturus (K2III), Hadar (B1III) and Aldebaran (K5III).
IV Subgiants
Stars which have begun evolving to giant or supergiant status. Some examples are Alnair (B7IV) and
Muphrid (G0IV). Note also Procyon which is entering this category and therefore is: F5IV-V.
V Dwarfs
All normal hydrogen-burning stars. Stars spend most of their lives in this category before evolving up
the scale. Class O and B stars in this category are actually very bright and luminous and generally
brighter than most Giant stars. Some examples are the Sun (G2V), Sirius (A1V), and Vega (A0V).
The M-K classification has been extended (IMA2 p. 225) to include subdwarf stars (VI or sd but not
universally to white dwarfs (D).
115
Other authors include a VII class for white dwarfs (aka wD or WD).
Other luminosity class extensions sometimes used include 0 [zero] and 1a-0 for hypergiants (e.g., Rho
Cassiopeiae is a G2Ia0e yellow hypergiant. Extension Iab signifies intermediate between Ia and Ib). Final
extension e means emission lines present.
Spectral peculiarity designations appended at the end, such as e, ep, k, m, n, p, s, var, v, !, etc., are
summarized here.
116

113
http://en.wikipedia.org/wiki/Hertzsprung%E2%80%93Russell_diagram
114
http://www.atlasoftheuniverse.com/hr.html
115
However, the table on IMA2 p. 228 shows D white dwarfs listed as part of the M-K luminosity class system
The radius of a star can be determined from its position in the H-R diagram by the Stefan-Boltzmann law. On
a logarithmically plotted H-R diagram, stars of the same radii fall along diagonal lines that run roughly
parallel to the main sequence.
The position of a main sequence star on the
main sequence (luminosity class V) is
determined primarily by its mass (see IMA2
App. G). The Main Sequence stars have a
density similar to water. The Sun, a G2V star,
has average = 1.4 g cm
-3
. Early-type stars
have lower average density: Sirius A, an A1V
star, has average 0.76 g cm
-3
= 0.54 ).
In contrast, the giant stars can have very low
average density. Belelgeuse, a supergiant
M2Iab star, has an average density of only
10
-8
.
Using the M-K system, astronomers can locate
a star on the H-R diagram, thereby estimate
its absolute magnitude, and then use its
apparent magnitude to estimate its distance
using the distance modulus. This technique
of estimating distance is loosely called by the
misnomer spectroscopic parallax.
The adjacent graph
117
shows the relationship
between spectral classes such as F0, color
index, luminosity classes (lines labeled with
Roman numerals), and absolute visual
magnitudes MV. Note that when the
horizontal axis is color index and the vertical
axis is absolute magnitude, the luminosity
classes and the boundaries separating the
spectral classes are all curving lines. (In some
diagrams, one is led to the mistaken
impression that the spectral classes are
demarcated by vertical lines.)

Another image of the H-R diagram, also including many non-main-sequence stars, follows:
118

116
Stellar Spectral peculiarity designations:
http://en.wikipedia.org/wiki/Stellar_classification#Spectral_peculiarities
http://www.skyandtelescope.com/howto/basics/3305876.html?showAll=y
117
118
image (modified by MCM) fr. http://stars.astro.illinois.edu/sow/hrd.html, fr. JB Kaler, The Cambridge
Encyclopedia of Stars, Cambridge U Press, Nov 2006

HR Diagram showing main-sequence and many other types of stars.
(Note that spectral classes depicted here are aligned vertically,
not always the case for H-R diagrams)
Stellar Atmospheres (Chapter 9, omitted)

This chapter has not yet been read in my astronomy courses and is not summarized here. Major topics
discussed in this chapter are:
The Radiation Field
Specific Intensity and Mean Intensity
Specific Energy Density
Specific Radiative Flux
Radiation Pressure

Stellar Opacity
Temperature and Local Thermodynamic Equilibrium
Definition of Opacity and Monochromatic Opacity (defined by dI/ds = -I)
Optical Depth
Sources of Opacity
Continuum Opacity and the H
ion
The Rosseland Mean Opacity

Radiative Transfer
Photon Emission Processes
The Random Walk
Limb Darkening (of a star at the edge)
The Radiation Pressure Gradient

The Transfer Equation
The Emission Coefficient j
The Special Case of Blackbody Radiation
The Assumption of Plane-Parallel Atmosphere
The Eddington Approximation
Limb Darkening Revisited

The Profiles of Spectral Lines
Equivalent Widths
Processes that Broaden Spectral Lines
The Voigt Profile (flux plotted against wavelength)
The Curve of Growth
Computer Modeling of Stellar Atmospheres

The Interiors of Stars (Chapter 10)

This chapter discusses stellar models and derives some of the needed formulas. It is only partially
summarized here. Stars are gravitationally confined thermonuclear reactors whose composition evolves as
energy is lost to radiation and neutrinos.
119

Basic Equations

Hydrostatic Equilibrium
The relevant formula for a spherically symmetrical star or other body is derived (IMA2 p. 287):

where P is pressure, r is radius from the center of the star, Mr is the mass contained within a sphere of radius
r, is the local density, and g is the local acceleration of gravity = GMr/r
2
. When applied using computer
modeling, a central solar pressure of nearly 2.34x10
16
N m
-2
is computed, equivalent to 2.3x10
11
Earth atm.

The Equation of Mass Conservation
Again for a spherically symmetrical star (IMA2 p. 288),

Pressure Equation of State
A state equation (equation of state EOS) is a relationship between state variablestemperature, pressure,
volume, entropy, enthalpy, internal energy, etc.whose values depend only on the current state of the system
and not the path by which the state was attained. The thermal gas pressure is given by

where Pg is the local ideal pressure of the gas, is local density, k is Boltzmann constant, T is temperature,
and is the mean molecular weight of the gas.
120
(For stars with ionized gases, this is actually the mean
particle weight, counting free electrons as well as ions as individual particles):

where is the average mass of a gas particle (including numbers of electrons in the denominator), and mH is
the mass of a hydrogen atom = 1.67 x 10
-27
kg. Clearly, the mean molecular weight depends on the state of
ionization. (IMA2 p. 288)

The Average Kinetic Energy Per Particle

This is given by

119
Woosley SE, et al, The evolution and explosion of massive stars, Rev. Mod. Phys. 74, 10151071 (2002)
120
http://www.vikdhillon.staff.shef.ac.uk/teaching/phy213/phy213_molecular.html
T
which reflects the three coordinate directions or degrees of freedom (IMA2 p. 294).

Radiation Pressure
The contribution to pressure attributable to photon radiation is given by (IMA2 p. 295)
c

where T is temp in Kelvin, =radiation constant = 4/c , = StefanBoltzmann constant = 5.6710
8
W m
2

K
4

Total Pressure
The pressure is the sum of the ideal gas pressure and the radiation pressure (IMA2 p. 295):

A solar model gives a central temperature of 1.57x10
7
K. At this temperature, radiation pressure is 1.53x10
13

N m
-2
, only 0.065% of gas pressure and therefore negligible. However, in larger stars radiation pressure plays
a major role. A third contributor to pressure, degeneracy pressure, is tiny for the Sun but becomes dominant
in degenerate stars.

Gravitation and the Kelvin-Helmholtz Timescale
One likely source of stellar energy is conversion during collapse of gravitational potential energy. A crude
calculation assuming a constant density throughout gives an estimated total gravitational energy of

where G is the gravitational constant, M is the total mass of the star, R is the current radius of the star.

The total mechanical energy E of the star by the virial theorem is

and this same amount of energy must have been radiated away or carried off during collapse (IMA2 p. 296)
The authors show that at constant luminosity L, the Sun would require only tKH = 10
7
years to emit this
much energy, where tKH represents the Kelvin-Helmholtz Timescale. This is inconsistent with the estimated
age of the solar system based on moon rocks etc., namely approximately 4x10
9
y. (However, gravitational
potential energy conversion can play a major role in other situations.)

Nuclear Reactions
I will give only a brief summary of this important and broad subject (see IMA2 p. 298ff)
For any atom, the number of nuclear protons (atomic number) is Z, the number of neutrons N, the total
nucleons A (called the mass number). The charge on the proton is +e, where -e is the charge on the electron.
Atomic particle masses are given by:

mp = 1.67262158 x 10
-27
kg = 1.0072765 u
mn = 1.67492716 x 10
-27
kg = 1.0086649 u
me = 9.10938188 x 10
-31
kg = 0.0005485799 u
where 1 u has E = mc
2
energy equivalence of 931.494013 MeV/c
2
(using m = E/c
2
).
The binding energy of a nucleus is the positive amount of mechanical energy required to disassemble it into
separate parts. A bound system has a lower potential energy than its constituent free partsthis is what
keeps the system together. Typically, the difference in mass (mass deficit) appears as energy released upon
the creation of a bound state (by fusion of nucleons). The reaction releases positive binding energy with the
same magnitude as the negative potential energy attained by the fused nucleus.
121
The binding energy of the
helium nucleus is 26.731 MeV.
As a rough calculation, conversion of 0.7% of the inner 10% of the Suns mass of H to He yields
Enuclear = 0.1 x 0.007 x Mc
2
= 1.3x10
44
J.
This amount of binding energy released at constant luminosity L for would last ~10
10
years, more than
sufficient to power the Sun over the expected age of 4x10
9
y.
In solar nuclear reactions at expected core temperatures of 1.57x10
7
K, the classical barrier of Coulomb
repulsion cannot be overcome to any significant degree. However, quantum mechanical tunneling based on
Heisenberg Uncertainty allows nucleons to overcome this Coulomb repulsion barrier and enter the deep
potential well leading to fusion. A rough calculation shows that 1.57x10
7
K is adequate to allow the protons
to approach the nucleus by a distance of one proton wavelength, giving the tunneling a realistic probability.
In calculating reaction probabilities, George Gamow (1904 - 1968) showed that a peak probability of reaction
exists (the Gamow peak) at a particular kinetic energy E of the colliding particles. If E is too low, the nuclear
Coulomb barrier cannot be overcome even by tunneling, whereas if E is too high, the e
-E/kT
Maxwell-
Boltzmann distribution term greatly reduces the likelihood that such energetic particles will be present (IMA2
p. 305 and here
122
).
Adjacent Graph: Gamow Peak
(graphed in red) for Fusion
Reaction Probability (fr. Fabrice
Fleurot
123
)
The Gamow peak occurs at
E0 = (bkT/2)
2/3
, where b is a
complicated term formed from the
reduced mass, Z1 and Z2, and
constants (IMA2 p. 304).

Nuclear reactions also often
exhibit resonances, typically sharp
localized increased reaction
probabilities for one or more
relatively narrow energy ranges...
Electron screening by electrons
liberated by ionization helps to
hide the nucleus and thus reduce
the effective Coulomb barrier and
improving the reaction rate (by up
to 10% to 50%)

121
http://en.wikipedia.org/wiki/Binding_energy
122
Graph and paraphrase from http://nu.phys.laurentian.ca/~fleurot/fusionrate/
123
ibid.
Nuclear reactions may also be expressed in terms of power laws in the form
e

where e
is the amount of energy liberated per kg of material, and
are the mass fractions of the two

particles, and and are exponents...

The Luminosity Gradient Equation
To compute the luminosity of a spherically symmetrical star, all of the energy generated by stellar material
must be considered:

where Lr is interior luminosity (energy production out to radius r), is the total energy released per kg per sec
by all nuclear reactions and by gravity (the latter may be negative)

Stellar Nucleosynthesis and Conservation Laws
Has Bethe is credited with original discovery in 1938 of details of conversion of H to He in stars as their
energy source, although Arthur Eddington (18821944) in 1920 had suggested that stars obtained their
energy from fusion.
A two-body collision is much more probable than a four-body collision, so that nucleosynthesis usually takes
place as a series or chain of two-body and one-body reactions.
In each reaction, there must be:
conservation of electric charge (adding positive and negative charges including those of antimatter),
conservation of number of nucleons (neutrons and protons), and
conservation of leptons (i.e., number of leptons minus number of antimatter leptons)
(leptons are electrons, positrons, neutrinos, antineutrinos, muons, taus, not photons)
Neutrinos have a tiny nonzero mass, and have such a low collision probability that almost all created deep
inside the Sun escape at the surface (IMA2 p. 309).
Atoms are represented by the following notation:

where A is total number of nucleons (atomic mass number), Z is number of protons (atomic number), and X is
the element determined by Z.

Proton-Proton Chains
These are one of the main chain of reactions that in sum convert H to He. The PPI branch, which is the most
common in the Sun (69%), has reactants and intermediates of ordinary hydrogen or protium (
), deuterium
(
) and helium-3 (
e) and emissions of positrons, neutrinos, and gamma rays (IMA2 p. 309).

e +e

The PPII branch of the PP chain (~31% on the Sun) has reactants and intermediates of helium-3 ( e
) ,
helium-4 e
, beryllium-7 e
, and lithium-7
, with emission of electrons, neutrinos, and gamma

rays (IMA2 p. 310).

The PPIII branch of the PP chain, which is quite rare on the Sun (<1%), has reactants with electron capture
and intermediates of protium
, beryllium-7 e
, boron-8
, and beryllium-8 e
, with emission of
positrons, neutrinos, and gamma rays (IMA2 p. 310).

At the Suns core temperature (estimated at 1.5 x 10
7
K), the energy generation rate is proportional to T
4
and
the PP chain fusion reactions predominate.

CNO Cycle
This independent cycle, formulated by Hans Bethe (1906 - 2005) in 1938, also converts H to He. The first
branch produces carbon-12 and helium-4, has reactants and intermediates of carbon-12 (
), protium
,
nitrogen-13 (
), carbon-13 (
), nitrogen-14 (
), oxygen-15 (
), nitrogen-15 (
), and Helium-4
(
e), with emissions of positrons, neutrinos, and gamma rays (IMA2 p. 311).

A second CNO Cycle branch is much less probable. It produces nitrogen-14 and helium-4 and involves
reactants and intermediates of carbon-12, protium, nitrogen-13, carbon-13, nitrogen-14, oxygen-15, and
nitrogen-15, with emissions of positrons, neutrinos, and gamma rays (IMA2 p. 311).

Comparing PP and CNO cycle reactions, the CNO cycle has a much stronger temperature dependence, with
the energy generation rate is proportional to about T
20
. Low mass stars (less massive than M) with cooler
cores have predominantly PP hydrogen burning, whereas more massive stars (at least somewhat more
massive than M) have H burning dominated by the CNO cycle. CNO cycle burning is seen when the
temperature has risen to around 15 x 10
6
K.

Transition to Helium Burning in the Core
When the stellar core H has been mostly consumed (burned up), the mean molecular weight of the gas has
increased if T and are unchanged. By the pressure equation of state, if T and are unchanged, the central
pressure drops, as given by

This fall in pressure this leads to a partial collapse of the core until the T and density rise to oppose further
collapse. When T and density have become sufficiently high, helium nuclei can begin to burn (fuse).
Helium burning takes place by the Triple Alpha Process, in which three helium-4 nuclei are fused to one
carbon-12, with an intermediate of beryllium-8 and emission of gamma rays. This reaction is strongly
temperature dependent, with the energy generation rate is proportional to T
41
. It is seen when the
temperature has risen to around T ~ 100 x 10
6
K
124

Carbon and Oxygen Burning and Nucleosynthesis of Higher Elements
These reactions can also take place in the high temperature environment of helium burning. Carbon-12 and
helium-4 combine to form oxygen-16, which combines with helium-4 to produce neon-20, with emission of
gamma rays. This reaction takes place with T around 15 x 10
6
K.

124
http://en.wikipedia.org/wiki/Triple-alpha_process
The rising Coulomb barrier makes nucleosynthesis of larger nuclear species less and less likely. But with
temperatures rising to the range of 600 x 10
6
K, it may be possible to synthesize oxygen-16, neon-20, sodium-
23, magnesium-23, and magnesium-24.
With temperatures around 1000 x 10
6
K, magnesium-24, silicon-28, phosphorus-31, silicon-31, and sulfur-
32 may be formed (IMA2 p. 313).
Some of these reactions are endothermic (require net input of energy), and are more likely to occur in stellar
cores than exothermic reactions.

Binding Energy Per Nucleon, Nuclear Stability, and Relative Abundances
The binding energy per nucleon is shown in this graph:
125

Here, the high binding energy per
nucleon (BEPN) of Iron-56 is shown,
the isotope of greatest nuclear
stability. Higher Z atoms have lower
BEPN, as do lower-Z elements.
Helium-4, carbon-12, and oxygen-16
are unusually stable for their
positions and are relatively
abundant. Tritium is much less
stable than deuterium (and is
radioactive). Initially after the Big
Bang, most atoms in the universe
are thought to have been H or He,
but nucleosynthesis within stars and
supernovas is thought to have
created higher-Z elements
throughout the universe.

Energy Transport and Thermodynamics

I will give only highlights of these topics.
Transport of energy from the core to the atmosphere takes place by radiation, convection, and conduction
(although conduction is insignificant in most stars during most of their lifetimes).

Radiative Temperature Gradient
The temperature gradient for radiative transport in a spherically symmetrical star is:
T

where the term is the total Rosseland
126
mean opacity (IMA2 p. 250), =radiation constant = 4/c, is the
local density, r is distance from center, Lr is interior luminosity (energy production out to radius r) (IMA2 p.
316).

125
http://en.wikipedia.org/wiki/File:Binding_energy_curve_-_common_isotopes.svg
126
after Svein Rosseland (1894 - 1985)
Pressure Scale Height H
P

This is a characteristic height of convective regions, and is given by

where P is the pressure, is the local density, and g is the local acceleration of gravity = GMr/r
2
. A typical
value for the Sun is ~ R/10.
Formulas for First Law of Thermodynamics (IMA2 p. 318), Specific Heats CV and CP = CV + nR, parameter
CP/CV, and the adiabatic gas law are omitted here.

Adiabatic Speed of Sound
This derives from the compressibility of a gas and its inertia (represented by density):
v
/

Adiabatic Temperature Gradient
This describes how the temperature in a gas bubble changes as it rises from convection and expands
adiabatically. It is given by:
T

where k = Boltzmann constant, CP/CV, etc. [see earlier this chapter]. If the actual temperature gradient is
steeper than this adiabatic gradient (a superadiabatic gradient), nearly all of the interior luminosity may be
carried in this region by convection. Thus, either radiation or convection often predominate in a particular
region, depending on the temperature gradient (IMA2 p. 322).

Criteria for Convection
The criteria derived for convection of a hot gas bubble are:
|
T
|
T

or equivalently,

T
<

where again CP/CV. For a monoatomic gas, =5/3 and convection occurs where the LHS of the 2nd
equation is < 2.5. If the LHS is > 2.5, the region is stable against convection.
Convection tends to occur when one or more of these apply:
(1) the stellar opacity is large (limiting radiative transport because an unachievably steep temp.
gradient would be required),
(2) a region exists where ionization is occurring, causing a large specific heat and a low adiabatic
temperature gradient, and/or
(3) the temperature dependence of the nuclear energy generation rate is large, causing a steep
radiative flux gradient and a large temperature gradient.
Items (1) and (2) can occur in the atmospheres of stars simultaneously, whereas (3) only occurs deep in stellar
interiors when CNO or triple-alpha processes are occurring (IMA2 p. 325).

Mixing-Length Theory of Superadiabatic Convection
The authors show that the temp gradient need be only slightly superadiabatic in the deep interior in order for
convection to carry most of the energy. The derivation introduces the mixing length l, which represents the
distance a hot bubble must rise before it thermalizes (its excess heat/temperature is dissipated to the
surrounding gas). HP is the pressure scale height (distance for pressure to drop to 1/e = 0.37 of the highest
or inner value)... The convective flux FC is derived. Mixing-Length Theory uses arbitrary empirical constants
but has been relatively successful in predicting the results of observations.
In an example calculation, a convective velocity for the Sun at r=0.714 is estimated at 50 m s
-1
, which is
10
-4
vs.(the local solar sound speed).

Stellar Model Building
Section discusses Entropy, Constitutive Relations, Boundary Conditions, the Vogt-Russell Theorem,
Numerical Modeling, Polytropic Models (special cases which allow analytical solutions), and Poissons
Equation. The highly idealized Lane-Emden Equation is derived (IMA2 p. 336).

The Main Sequence

As a result of H burning in the stellar core, the Vogt-Russell Theorem requires that such a change of the
composition or mass of a star leads to adjustments in the effective temperature and luminosity. In other
words, what can be observed arising from the stars surface reflects the state of the star deep within. As
stellar mass increases, the central P and T increase.

Lower Mass Stars and Lower Main Sequence Mass Limit
As initial star mass decreases, a lower limit is reached at which point the central hydrogen-burning nuclear
reactions (if ever present) are insufficient to maintain a sufficiently high temperature to stabilize the star
against gravitational contraction (see later discussions). Such contracted stars are no longer on the main
sequence. This lower mass limit is about 0.072 M (for Population I having high Z)it is somewhat higher,
0.09, for Population II stars with low metallicity Z0). (IMA2 p. 341)
For lower mass main sequence stars (< 1.2 M), radiation is adequate in the core to transport the energy
released and there is no core convection. However, convection forms near the surface due to the fall of
effective temperature with decreasing mass, causing increasing opacity arises partly from H ionization, which
makes convection more efficient. The inner limit of convection descends closer to the center with decreasing
mass, until at 0.2 M and smaller, the entire star is convective. (IMA2 p. 343)

Higher Mass Stars and Upper Main Sequence Mass Limit (Eddington Luminosity Limit)
Larger stars have higher luminosities and at an upper limit in size and luminosity, the radiation pressure can
dominate over the gas pressure in certain regions, often in the outer layers. Arthur Eddington (1882 - 1944)
gave us the Eddington limit on luminosity:

where M is the stars mass, and the kappa term is the total Rosseland mean opacity (for which larger values
signify greater opacity and more rapid intensity attenuation). This is the maximum radiative luminosity that
a star can have and still remain in hydrostatic equilibrium. Beyond this luminosity, the envelopes of massive
stars become more loosely bound, and such larger stars suffer large mass loss from being blown off as
stellar wind (IMA2 p. 343 and here
127
). The Eddington limit for pure H II is about 3.2 x 10
4
(M/M) L.
Eddingtons formula has been updated, does not require spherical symmetry, and is now applied to novae,
gamma ray burst, and accretion disks as well as to stars. Stars of about 90 M are near this limit. High
mass main sequence stars (> 90 M) exhibit rapid mass loss and often oscillatory or variable nuclear energy
generation.
The term Eddington accretion rate (an analogous limit) is applied to accretion disks, see Chapter 18.
For larger mass main sequence stars, convection is dominant at their cores because the CNO-cycle burning is
strongly temperature dependent producing a strong temp gradient, plus radiation is not efficient enough to
transport the rate of energy release. Outside of the hydrogen burning core, convection ceases as radiation
can handle the transport. (IMA2 p. 343) Stars less than 10 M may also have an outer convection zone.
128

Range of Luminosities and Lifetimes
Over the range of masses of main sequence stars, luminosities range from a lower limit of about 5x10
-4
L
(Teff 1700K) to an upper limit of about 10
6
L (Teff 53,000K), spanning nine orders of magnitude. The
much greater luminosities of massive stars are associated with correspondingly much shorter lifetimes. Of
course, the substantial differences in effective temperatures are associated with characteristic changes in the
stellar spectra.

127
http://en.wikipedia.org/wiki/Eddington_luminosity
128
http://en.wikipedia.org/wiki/Convection_zone
The Sun (Chapter 11)

The Solar Spectrum (starting in near UV, including visible light (400 to 700 nm), and extending to IR)
129

The blackbody peak for the Suns effective temperature of 5777 K is 502 nm by Wiens Displacement Law.

Note: I have also presented substantial information about the Sun in a Space Physics summary here.
130

The Solar Interior

Our star, the Sun, is divided into an atmosphere (photosphere, chromosphere, transition region, and corona)
and an interior (radiative and convection zones). Going outward, the base of the atmosphere may be taken to
be the region in which optical depth has fallen to less than unity. (See below for a more exact definition of the
base of the photosphere.) The optical depth at wavelength for a path through a stellar atmosphere is given
by

For an optical depth 1, the intensity of light at wavelength decreases to e
-1
= 1/e = 0.368 of the initial
intensity value. At an optical depth = 2, a reduction to 1/e
2
= 0.135 of the initial intensity value occurs.

Evolutionary History of the Sun
The Sun is spectral class G2 luminosity class V star. The surface (atmospheric) composition is X = 0.74, Y =
0.24, and Z = 0.02 (where XYZ are the mass fractions of H, He, and metals, resp.) Comparison with the

129
http://rredc.nrel.gov/solar/spectra/am1.5/ASTMG173/ASTMG173.html.
Reference solar spectral irradiance data and graphs are available here, including the graphable Excel data
shown above. Graph created by MCM.
130
http://www.mcgoodwin.net/pages/spacephysics_ess471.pdf
unchanging moon and meteoritic rocks places the Suns age is estimated at 4.57 Ga (Ga = Gy= 10
9
Julian
years, each of which is 31557600 SI seconds in length).
131

According to stellar models,
132
during the Suns evolution as a main sequence star over 4.57 Ga,

the luminosity L of the Sun has steadily risen from an initial value of 0.677 L to 1 L at the present. It will
continue to rise at least to 8 Ga.
the radius of the Sun R has steadily risen from an initial value of 0.869 R to 1 R currently. It will
continue to rise at least to 8 Ga.
the effective temperature of the Sun teff has steadily risen from an initial value of 5620 K to 5777K currently.
The model shows that it will rise further and attain a peak near 5800 K at around 6.5 Ga, then drop to 5780
K by 8 Ga.
The impact of these changes on Earth depends on complex factors including Earth atmospheric convection,
Earth atmospheric composition, and Earth atmospheric and ground albedo.

The Present Interior of the Sun
The textbook presents considerably more detail than summarized here, including useful graphs of
distributions of P, T, density, etc.
According to the 2001 models of Bahcall, Pinsonneault, & Basu, the core temperature is 1.57 x 10
7
K and this
drops steadily outward. The core pressure is 2.342 x 10
16
N m
-2
, and this falls off more rapidly with distance
than does T. The core density 1.527 x 10
5
kg m
-3
and this also falls off rapidly similar to P. The current core
mass fractions are X = 0.3397 (H has decreased from initial 0.71) and Y = 0.6405 (He has increased from
0.27). Near the surface, gravitational settling has decreased Y by 0.03 and enriched X by 0.03.
The base of the present day solar convection zone (the tachocline
133
) is at 0.714 R. Below this level, the Sun
is purely radiative. Convection arises when the temperature gradient is superadiabatic,
|
T
|
T

and for reasons previously discussed (see Chapter 10). (For more massive stars, convection begins in the
center of the star, Chapter 13).
The peak rate of energy production (luminosity) from fusion is seen at about 0.1 R. (This result partly
reflects the amount of mass contained in a tiny shell at the center vs. a much greater mass in a shell of
identical width at 0.1 R, and also reflects the maximal depletion of H in the center.) Luminosity does not
increase beyond about 0.3 R because there is the temperature is too low for effective fusion.
90% of the mass of the sun is enclosed at about 0.5 R.

Differential Rotation of the Interior
The rotation rate of the interior of the Sun varies with depth. Deep to the tachocline (in the radiation zone),
the rotation angular velocity is essentially constant with changing r/R. However, in the convection zone of
the interior lying above the tachocline, the rate of rotation varies with both latitude and depth.
134
The shear at

131
Bouvier A, Meenakshi W, "The age of the solar system redefined by the oldest Pb[207]-Pb[206] age of a
meteoritic inclusion". Nature Geoscience, Published online: 22 August 2010
http://www.nature.com/ngeo/journal/v3/n9/full/ngeo941.html
132
Bahcall HN, Pinsonneault MH, & Basu S, Solar Models: Current Epoch and Time Dependences, Neutrinos,
and Helioseismological Properties, The Astrophysical Journal, 555:990-1012, 2001 July 10,
http://iopscience.iop.org/0004-637X/555/2/990/fulltext/52957.text.html and IMA2 p. 350
133
Tachocline:
http://solarscience.msfc.nasa.gov/interior.shtml
http://en.wikipedia.org/wiki/Tachocline
134
http://en.wikipedia.org/wiki/Differential_rotation
the tachocline is thought to generate electric currents and play a role in generation of the magnetic field (IMA2
p. 364).

The Solar Neutrino Problem
Experiments by Raymond Davis and John Bahcall (using data gathered deep in the Homestake Gold Mine),
and subsequent experiments by Japans Super-Kamiokande observatory
135
and SAGE and GALLEX
experiments, measured rates of neutrinos generated in the Suns core and captured in the experimental
system. Davis expressed reaction rates as solar neutrino units (SNUs), defined as 10
-36
reactions per second
per target atom (Chlorine-37 conversion to radioactive Argon-37, in Daviss case). They consistently found
insufficient SNUs according to the then prevailing models. The explanation came with neutrino oscillations in
the Mikheyev-Smirnov-Wolfenstein effect, changing our understanding of particle physics.
136

The Solar Atmosphere

The Suns apparently sharp edge (in visible light) is actually formed in a region 600 km thick (0.09% of R)
over which the atmosphere changes rapidly from optically thick (small optical depth) to optically thin (larger
optical depth).

Photosphere
Because of the wavelength-dependence of optical depth, the base of the photosphere is defined relative to the
level where the optical depth = 500 nm = 1. That is, the of the atmosphere at a wavelength of 500 nm is 1,
and the intensity of a light ray is reduced to 1/e = 0.368. The base of the photosphere is arbitrarily taken to
be 100 km deeper than the 500 nm = 1 level. At this only slightly deeper level, the optical depth is much
greater (500 nm = 23.6), almost no light can penetrate by radiation, and the temperature is about 9400 K.
The photosphere is where most of the visible (optical) photons we observe originate (very few come from below
the base as defined). Temperature in the photosphere decreases with increasing radius from the Sun center,
until one reaches a minimum value of about 4400 K at the outer limit of the photosphere. This region of
minimum temperature is about 525 km above the 500 nm = 1 level or 525 km above the photosphere base as
defined above. Thus, the entire photosphere is about 625 km thick.
The effective temperature of the photosphere, which we measure to be about 5777 K, derives from the average
level from which photons that reach us have arisen. In the Eddington approximation, this proves to be where
= 2/3 and this corresponds to Teff = T = 2/3 = 5777 K, part way between the extremes of 9400 K and 4400 K.
The visible and IR solar spectrum resembles that of a blackbody continuum, consistent with a source of
opacity for these bands. The opacity is thought to arise from the presence of H
ions (a hydrogen atom with

extra electron)although they are rare, they are much more likely than H I to contribute to continuum
radiation. Different parts of the solar spectrum may arise from different levels in the photosphere.
Absorption lines are also produced in the photosphere. The darkest central part of the lines are produced in
the cooler outer regions of the photosphere. Off center from the dark peak, the lesser absorption arises from
deeper hotter levels (MCM: presumably because of higher velocities and greater Doppler spread of
wavelengths).
Granules: I have discussed solar granulations and other Sun topics here.
137
Granulations arise from
convecting cells spanning about 700 km and lasting 5 to 10 minutes. Spectra obtained across these
granulations show wiggles of the lines from Doppler shifts due to the rising (blueshifted, brighter, and more

135
http://www-sk.icrr.u-tokyo.ac.jp/sk/index-e.html
136
Neutrino oscillations:
http://www.nobelprize.org/nobel_prizes/physics/articles/bahcall/
http://en.wikipedia.org/wiki/Solar_neutrino_problem
137
central) and descending (redshifted, darker, and more peripheral) portions of gas. Granules are small (about
1000 km across) cellular features that cover the entire Sun except for those areas covered by sunspots. These
features are the tops of convection cells where hot fluid [gas] rises up from the interior in the bright areas,
spreads out across the surface, cools and then sinks inward along the dark lanes. Individual granules last for
only about 20 minutes. The granulation pattern is continually evolving as old granules are pushed aside by
newly emerging ones... The flow within the granules can reach supersonic speeds of more than 7 km/s ... and
produce sonic booms and other noise that generates waves on the Sun's surface.
138

Faculae: Faculae are bright areas [in the photosphere] that are usually most easily seen near the limb, or
edge, of the solar disk. These are also magnetic areas but the magnetic field is concentrated in much smaller
bundles than in sunspots. While the sunspots tend to make the Sun look darker, the faculae make it look
brighter. During a sunspot cycle the faculae actually win out over the sunspots and make the Sun appear
slightly (about 0.1%) brighter at sunspot maximum that at sunspot minimum.
139

Differential Rotation of the Photosphere
The rotation rate (angular velocity) of the gas of the photosphere also varies with latitude, from a period of
about 25 days near the equator to about 36 days near the poles. I have discussed solar differential rotation in
relation to magnetic fields and sunspots here.
140

Chromosphere
This sphere of color lies just above the photosphere, starting therefore at about 525 km above the 500 nm =
1 level and extending about 1600 km to 2100 km above the 500 nm = 1 level. The gas density drops by a
factor of 10
4
over this distance, and the temperature gradually rises from 4400 K at the bottom to about
10,000 K at the top. This region imparts lines to the solar spectrum that are not produced in the
photosphere, particularly H Balmer lines and those of He II, Fe II, Si II, Cr II, and Ca II (especially the H 396.8
nm line and K 393.3 lines). The lines added include some emission lines arising in the outer hot zone,
according to Kirchhoffs second law of spectroscopy.
The chromosphere is not usually visible, as it is overwhelmed by the photosphere, but can be seen in a solar
eclipse as a reddish region. In addition, the flash spectrum as the Suns upper limb just rises or sets appears
as red from the Balmer H 656.3 nm emission line arising in the chromosphere.
141
Imaging using the Balmer
H emission line reveals supergranulation on a scale of 30,000 km in the solar atmosphere, apparently
reflecting underlying convective phenomena (IMA2 p. 365 and here
142
).
The chromosphere may also be imaged using the 393.4 nm emission line of Ca II [Ca II K line]. When the Sun
is viewed through a spectrograph or a filter that isolates the H-alpha emission, a wealth of new features can
be seen. These features include the chromospheric network of magnetic field elements, bright plage around
sunspots, dark filaments across the disk and prominences above the limb.
143

Spicules: Spicules are vertically-oriented filaments of rapidly ascending gas (15 km s
-1
) extending as much as
10,000 km or more above the usual level of the outer chromosphere into the corona.
144
These are relatively
small but widespread eruptions, and thus they eject substantial material into the corona. These are also best
observed in the H emission line. They last but a few minutes but in the process eject material off of the
surface and outward into the hot corona at speeds of 20 to 30 km/s.
145

138
http://solarscience.msfc.nasa.gov/feature1.shtml
139
ibid
140
ibid.
141
http://www.phys.vt.edu/~heremans/Astrolab1156/Readings/flash.html
142
Solar Supegranulations:
http://solarphysics.livingreviews.org/Articles/lrsp-2010-2/download/lrsp-2010-2Color.pdf
http://solarscience.msfc.nasa.gov/feature1.shtml (this author lists supergranulation under photospheric
phenomena rather than chromospheric)
143
http://solarscience.msfc.nasa.gov/chromos.shtml
144
http://history.nasa.gov/SP-402/p2.htm
145
Plage: A plage (from Fr. beach) is a bright region in the chromosphere of the Sun, typically found in regions
of the chromosphere near sunspots. The plage regions map closely to the faculae in the photosphere below...
146
Plage ... are bright patches [in the chromosphere] surrounding sunspots that are best seen in H-alpha.
Plage are also associated with concentrations of magnetic fields and form a part of the network of bright
emissions that characterize the chromosphere.
147

Filaments: Filaments are dark, thread-like [chromospheric] features seen in the red light of hydrogen (H-
alpha). These are dense, somewhat cooler, clouds of material that are suspended above the solar surface by
loops of magnetic field.
148

Prominences: Prominences are dense clouds of material suspended above the surface of the Sun by loops of
magnetic field. Prominences and filaments are actually the same things except that prominences are seen
projecting out above the limb, or edge, of the Sun. Both filaments and prominences can remain in a quiet or
quiescent state for days or weeks. However, as the magnetic loops that support them slowly change,
filaments and prominences can erupt and rise off of the Sun over the course of a few minutes or hours.
149

A prominence is a large, bright feature extending outward from the Sun's surface, often in a loop shape.
Prominences are anchored to the Sun's surface in the photosphere, and extend outwards into the Sun's
corona. While the corona consists of extremely hot ionized gases, known as plasma, which do not emit much
visible light, prominences contain much cooler plasma, similar in composition to that of the chromosphere...
A typical prominence extends over many thousands of kilometers; the largest on record was observed by the
Solar and Heliospheric Observatory (SOHO) in 2010 and is estimated at over 700,000 kilometers... long
roughly the radius of the Sun... A prominence forms over timescales of about a day, and stable prominences
may persist in the corona for several months. Some prominences break apart and give rise to coronal mass
ejections.
150

The Transition Region
The temperature rises rapidly at the top of the chromosphere in the transition region, attaining 10
5
K in only
100 km and about 10
6
K in 200 km. It can be seen in the Ly- emission line, the CIII 97.7 nm line, and other
UV emission lines seen at high temperatures (IMA2 p. 365). The transition region does not occur at a well-
defined altitude. Rather, it forms a kind of nimbus around chromospheric features such as spicules and
filaments [that extend into the corona], and is in constant, chaotic motion. The transition region is not easily
visible from Earth's surface, but is readily observable from space by instruments sensitive to the extreme
ultraviolet portion of the spectrum.
151

The transition region is a thin and very irregular layer of the Sun's atmosphere that separates the hot corona
from the much cooler chromosphere. Heat flows down from the corona into the chromosphere and in the
process produces this thin region where the temperature changes rapidly from 1,000,000C ... down to about
20,000C... Hydrogen is ionized ... at these temperatures and is therefore difficult to see. Instead of
hydrogen, the light emitted by the transition region is dominated by such ions as C IV, O IV, and Si IV...
These ions emit light in the ultraviolet region of the solar spectrum that is only accessible from space.
152

The Corona
The corona (crown), a plasma atmosphere located above the transition region, becomes visible only in solar
eclipses. Its energy output is 10
-6
that of the photosphere. It is extremely tenuous, having a particle number
density of 10
15
m
-3
at the base, thinning to 10
15
m
-3
in the solar wind arriving at the Earth. It is not in local
thermodynamic equilibrium, so temperatures are not strictly definable. However, the presence of highly

146
http://en.wikipedia.org/wiki/Plage_%28astronomy%29
147
148
ibid
149
ibid
150
http://en.wikipedia.org/wiki/Solar_prominence
151
152
http://solarscience.msfc.nasa.gov/t_region.shtml
ionized iron (Fe-XIV) lines accompanied by substantial Doppler line broadening indicates a plasma
temperature in excess of 1 to 2 x10
6
K.
The corona is the extended outer atmosphere of the Sun, which is much larger in volume than the Sun itself.
The corona continuously expands into space forming the solar wind, which fills all the Solar System. The low
corona, near the surface of the Sun, has a particle density around 10
15
10
16
m
3
. The average temperature of
the corona and solar wind is about 1,000,0002,000,000 K; however, in the hottest regions it is 8,000,000
20,000,000 K. While no complete theory yet exists to account for the temperature of the corona, at least some
of its heat is known to be from magnetic reconnection.
153

Coronal Light Components: Coronal radiation is subdivided into three partly coextensive components:
the K coronaKontinuierlich=continual white light arising between 1 and 2.3 R from scattering of sunlight
off free electrons,
the F coronanamed after Fraunhofer, but describing scattering from dust grains beyond 2.3 R, a
phenomenon which merges with the zodiacal light (emitted from the zodiacal cloud of dust), and
the E coronaEmission lines from highly ionized atoms, including forbidden transitions from metastable
states observable due to low number densities and the rarity of collisions.
Radio Emission: The corona is a source of radio-wavelength non-blackbody radiation, which arises from free-
free electron transitions (aka bremsstrahlung or braking radiation arising from electronion near-collisions).
Wavelengths of 1 to 20 cm are observed from the outer corona (mixed in with synchrotron radiation also
arising in the corona). (IMA2 p. 369 and here
154
)
X-ray Emission: The corona is also a source of X-rays, again not arising as blackbody radiation but rather
from multiply-ionized elements such as iron, oxygen, and neon. Wavelengths illustrated in IMA2 are 14 - 17
nm. X-ray emission is not uniform across the Sun, but arises in active hot regions. Weaker emissions arise
from darker and cooler coronal holes... (IMA2 p. 370) The corona shines brightly in x-rays because of its
high temperature. On the other hand, the "cool" solar photosphere emits very few x-rays. This allows us to
view the corona across the disk of the Sun when we observe the Sun in X-rays... In the early 70's Skylab
carried an x-ray telescope that revealed coronal holes and coronal bright points for the first time. During the
last decade, Yohkoh provided a wealth of information and images on the sun's corona. Today we have the
SOHO and TRACE [Transition Region and Coronal Explorer] satellites obtaining new and exciting
observations of the Sun's corona, its features, and its dynamic character.
155

Helmet streamers: Helmet streamers are large cap-like coronal structures with long pointed peaks that
usually overlie sunspots and active regions. We often find a prominence or filament lying at the base of these
structures. Helmet streamers are formed by a network of magnetic loops that connect the sunspots in active
regions and help suspend the prominence material above the solar surface. The closed magnetic field lines
trap the electrically charged coronal gases to form these relatively dense structures. The pointed peaks are
formed by the action of the solar wind blowing away from the Sun in the spaces between the streamers.
156

Coronal loops: Coronal loops are found around sunspots and in active regions. These structures are
associated with the closed magnetic field lines that connect magnetic regions on the solar surface. Many
coronal loops last for days or weeks but most change quite rapidly... Some loops, however, are associated
with solar flares and are visible for much shorter periods. These loops contain denser material than their
surroundings. The three-dimensional structure and the dynamics of these loops is an area of active
research.
157

Coronal Holes: Coronal holes are regions where the corona is dark. These features were discovered when X-
ray telescopes were first flown above the earth's atmosphere to reveal the structure of the corona across the
solar disc. Coronal holes are associated with "open" magnetic field lines and are often found at the Sun's
poles. The high-speed solar wind is known to originate in coronal holes.
158

153
154
http://www.cv.nrao.edu/course/astr534/FreeFreeEmission.html
155
http://solarscience.msfc.nasa.gov/corona.shtml
156
157
ibid.
158
ibid.

Model of chromospheric and coronal spicules,
and solar wind interacting with Earth magnetosphere (image from NASA)
159

Solar Flares
A solar flare is a sudden brightening observed over the Sun surface or the solar limb, which is interpreted as
a large energy release of up to 6 10
25
joules of energy (about a sixth of the total energy output of the Sun
each second)... The flare ejects clouds of electrons, ions, and atoms through the corona into space. These
clouds typically reach Earth a day or two after the event. The term is also used to refer to similar phenomena
in other stars, where the term stellar flare applies... The same energy releases may produce coronal mass
ejections (CME), although the relation between CMEs and flares is still not well established.
160

In a matter of just a few minutes [solar flares] heat material to many millions of degrees and release as much
energy as a billion megatons of TNT. They occur near sunspots, usually along the dividing line (neutral line)
between areas of oppositely directed magnetic fields... Flares release energy in many forms - ... Gamma rays
and X-rays, energetic particles (protons and electrons), and mass flows. Flares are characterized by their ...
X-ray flux. The biggest flares are X-Class flares [and have peak flux at 100 - 800 pm of >10
-4
W m
-2
]. M-Class
flares have [10
-5
10
-4
W m
-2
], and C-Class flares have [10
-6
10
-5
W m
-2
]... The key to understanding and
predicting solar flares is the structure of the magnetic field around sunspots. If this structure becomes
twisted and sheared then magnetic field lines can cross and reconnect with the explosive release of energy...
The flare (the bright area [see image]) lies along a section of a neutral line where the magnetic field is twisted

159
http://solarscience.msfc.nasa.gov/images/sunearth_lg.gif
160
http://en.wikipedia.org/wiki/Solar_flare
(or sheared) to point along the neutral line instead of across it. We have found that this shear is a key
ingredient in the production of solar flares.
161

Coronal Mass Ejections (CMEs)
A coronal mass ejection (CME) is a massive burst of solar wind, other light isotope plasma, and magnetic
fields rising above the solar corona or being released into space... Coronal mass ejections are often associated
with other forms of solar activity, most notably solar flares, but a causal relationship has not been
established. Most ejections originate from active regions on Sun's surface, such as groupings of sunspots
associated with frequent flares. Near solar maxima the Sun produces about three CMEs every day, whereas
near solar minima there is about one CME every five days.
162

Comparison of Major Solar Outbursts

Phenomenon Onset & Duration Energy &/or Size Location Comments
Solar Flare Explosively sudden
lasting milliseconds to
> 1 hr
Up to 6x10
25
J as
UV, X-ray, gamma
ray & all other EM
rad.; protons; up
to 100,000 km
affect all layers Categorize by X-
ray flux; may arise
from magnetic
reconnection or
sigmoids
163
of
energy
Solar
Prominence
1 day onset; lasts days
to months; stable
and/or eruptive
up to 1 R
arise in photosphere,
affect all layers
~ Loop shape,
some give rise to
CMEs
Coronal Mass
Ejection
164

onset over hours released matter
averages 510
13
kg
at 20 to 3200 km/s
as hot plasma + EM
radiation
usu. arise in active
sunspot regions,
some associated with
eruptive
prominences; cross
all layers
some but not all
assoc. w large
explosive solar
flares; cause
auroras; may arise
from magnetic
reconnection

Miscellaneous Solar Topics

Omitted Solar Topics
I have omitted summarization of the following topics:

Alfvn waves (transverse MHD waves) and their speed ( 10 m s
-1
)
165

Auroras
Heliopause and heliosheath
Helioseismology

161
Solar Flares:
http://solarscience.msfc.nasa.gov/flares.shtml
http://en.wikipedia.org/wiki/Solar_flare (quantitation of flux)
162
http://en.wikipedia.org/wiki/Coronal_mass_ejections
163
http://www.astronomynow.com/090420energyribbons.html
164
http://www-istp.gsfc.nasa.gov/istp/nicky/cme-chase.html
165
http://en.wikipedia.org/wiki/Alfv%C3%A9n_wave
Lorentz force (F = q (E + v B) (which follows the RH rule where v=index, B=middle finger, F=thumb)
Magnetic dynamo model of the solar cycle
166

Magnetic energy density
Magnetic pressure gradient
Magnetic reconnections
167

Magnetohydrodynamics MHD
Parker Solar Wind model (Heliospheric Current Sheet, Parker Spiral, Garden Hose Angle,
and Ballerina Skirt)
Solar cosmic rays
Solar magnetic fields
Solar wind and its Fast and Slow components
168

as these have been discussed in another course summary I wrote
169
or will be the subjects of more intensive
future courses I hope to take.

Solar Rotation Slowing
Stars like the Sun are observed to be slowing down in rotation rate. Late main-sequence stars with convective
envelopes have much slower rotation rates than stars on the upper end of the main sequence. This may be
due to transfer of angular momentum to solar wind.
170

Solar Cycle and Miscellaneous Variable Solar Activity

A brief summary follows, see also my other summary.
171

Sunspots
Sunspots, first observed by Galileo,
follow a ~11-year cycle from peak
activity (measured by sunspot
number
172
) to peak activity. They
tend to arise first at higher
latitudes (30-40) and gradually
extend toward (but not on) the
equator in the same cyclical
periodicity, producing the familiar
Butterfly diagram (upper right,
here from NASA
173
):
The temperatures of sunspot
umbras (the darkest central parts,
as large as 30000 km) are 3,900 -
4,100 K, whereas the surrounding
filamentary penumbras are
somewhat warmer . Compared to

166
http://solarscience.msfc.nasa.gov/dynamo.shtml
167
http://en.wikipedia.org/wiki/Magnetic_reconnection
168
http://solarscience.msfc.nasa.gov/SolarWind.shtml
169
170
IMA2 p. 380 and http://en.wikipedia.org/wiki/Stellar_rotation
171
172
Boulder and International Sunspot Numbers: http://spaceweather.com/glossary/sunspotnumber.html
173
NASA image from http://solarscience.msfc.nasa.gov/images/bfly.gif
the surrounding photosphere at about 5,777 K, the cooler sunspots have bolometric flux of about 1/5, and
are visible as dark spots. Despite their apparent darkness, even the umbra would be quite bright if viewed
without a darkening filter.
Using spectral Zeeman splitting (such as with the Fe I 525.02 nm line) and polarization data, the strongest
magnetic fields are observed in the umbras, where they are vertically oriented) whereas in the penumbras are
found somewhat weaker and more obliquely or horizontally oriented fields. The extra magnetic pressure in a
sunspot keeps it from sinking in (IMA2 p. 385).
Sunspots often occur in groups. The lead sunspot of a group has the same magnetic polarity in a solar
hemisphere during an 11-year sunspot cycle. In the other hemisphere or during the next 11-year cycle in the
same hemisphere, the polarity is reversed. The overall Sun polarity reverses during the sunspot minimum, so
that a return to the same polarity and peak sunspot activity takes about 22 years.
The cycling of sunspot activity is accompanied by cyclical variation in the overall solar luminosity of about
0.1% (IMA2 p. 383 and here
174
). It is hypothesized that frozen-in local magnetic fields associated with
sunspots inhibit convection bubbles at the sunspot leading to local cooling. (Note however that somewhat
paradoxically, the overall solar luminosity is maximal when sunspots are maximal).
The Maunder Minimum (c. 1645 - 1715) was a prolonged period of quiescent solar activity and sunspots,
colder Earth temperatures (Little Ice Age), and presumably lower solar luminosity. (Evidence to that effect is
deduced from carbon-14 and beryllium-10 abundances, isotopes which are affected by cosmic rays, in turn
affected by the Suns magnetic field).
175
John Eddy, who popularized the Maunder Minimum, suggests there
may in fact be a longer period variation in solar activity and luminosity that may last for centuries.

Total Solar Irradiance TSI
TSI is the spatially and spectrally integrated solar
radiation incident at the top of the Earth's atmosphere
(i.e., extraterrestrial and therefore unaffected by
Earths atmosphere). It is usually averaged over at
least a day to avoid the transient effects of sunspots,
etc., and is often expressed adjusted to 1 A.U. distance
from the Sun. The TSI used to be called the Solar
Constant but it is obviously not constant:

Historical Reconstruction of Total Solar Irradiance
1610 2011 (graph to right):
Based on Wang, Lean, and Sheeley, The Astrophysical
Journal, 625:522, 2005 May 20, and extended with
SORCE data... Data for this reconstruction was
computed by G. Kopp using TIM V.12 data Jan 19,
2012.
176
SORCE = Solar Radiation and Climate
Experiment, a NASA-sponsored satellite mission orbiting
at 645 km. (See also above how the luminosity of the
Sun has risen over its much longer entire 4.57 Ga
history from an initial 0.677 L)

SORCE Daily-Average TSI 2003 2012 (graph to right)
Data from Laboratory for Atmospheric and Space Physics
(LASP)
177
, adjusted to 1 A.U.

174
http://en.wikipedia.org/wiki/Solar_variation
175
http://en.wikipedia.org/wiki/Maunder_Minimum
176
http://lasp.colorado.edu/sorce/data/tsi_data.htm
177
http://lasp.colorado.edu/lisird/sorce/sorce_tsi/index.html
Using their reported daily average data, the average value of TSI for the period shown (2/25/2003
3/25/2012) is 1361.05 W m
-2
at the upper limit of the Earths atmosphere adjusted to a distance of 1 A.U.
The SORCE average true Earth irradiance (i.e., still extraterrestrial but not corrected to a distance of 1 A.U.)
for the same period is nearly the same, namely 1361.24 W m
-2
.
Other Phenomena
Many of the variable solar phenomena are mentioned in connection with the atmospheric layers.
Comet tails extend beyond the coma surrounding the nucleus and become visible as comets near the Sun and
are illuminated and heated by it, inducing vaporization. The dust reflects sunlight directly and the gases glow
from ionization. The streams of dust and gas each form their own distinct tail component. The tail of dust is
left behind in the comet's orbit in such a manner that it often forms a curving tail called the antitail. The ion
tail, made of gases and sometimes called the true tail, always points directly away from the Sun. The ionized
gas is more strongly affected by the solar wind than is dust, and follows the Suns magnetic field lines rather
than an orbital trajectory From the Earth, the tails may appear to point in opposite directions.
178

Similar Behavior in Other Stars
Other stars also exhibit magnetic activity cycles (including Zeeman splitting), and slowing of overall rotation.
Flare stars, which are main-sequence spectral class M, exhibit rapid changes in brightness possibly due to
flares. Star spots have been inferred by their effect on luminosity, especially notable in RS Canum
Venaticorum and BY Draconis stars (IMA2 p. 393 and these
179
).

178
Comet Tails:
modified from http://en.wikipedia.org/wiki/Comet_tail
http://www.universetoday.com/40820/comet-tail/
179
Starspots:
http://www.livingreviews.org/lrsp-2005-8
http://go.owu.edu/~physics/StudentResearch/2003/BethCademartori/index.html
The Interstellar Medium; Protostar and Early Star Formation (Chapter 12)

Interstellar Dust and Gas

Star evolution is a cyclical process, in which stars are born from gas and dust in the interstellar medium
(ISM). Stellar winds, planetary nebulae, and explosive supernovae etc. return some of this material to the
ISM. We must consider turbulent gas motions, shocks, magnetic fields, magnetohydrodynamics,
thermodynamics, quantum mechanics, chemistry, radiative transfer of the ISM to properly study star
evolution.
Dust forms probably mostly from coagulation within ISM molecular clouds, but also from supernova
explosions and stellar winds and in the envelopes of cool stars (IMA2 p. 411).
Dust in the ISM contributes to extinction/attenuation, reddening of transmitted light, reprocessing of UV and
optical light to IR, scattering of light, and locking up of metals (ZXI lecture 2). These topics are discussed
below or elsewhere in this document.
Interstellar Light Extinction (Attenuation) A

This is an astronomical term describing the effects of absorption and scattering of electromagnetic radiation
by matter (dust and gas) between an emitting astronomical object and the observer.
180
It is often expressed as
a positive magnitude of wavelength-dependent extinction A where apparent magnitude is given by:
m = M + 5 log10(d) - 5 + A
The extinction measured in magnitude changes in the visual band and occurring from dust measured is
expressed as AV.
Extinction may be said to account for the dark bands in the Milky Way, the Horsehead Nebula, and the Bok
globule B68.
181

For wavelength-dependent optical depth defined in
,
e
, the authors derive A = 1.086, so that the

change in magnitude due to extinction is approximately equal to the optical depth . Moreover, = Nd,
where = scattering cross section for the dust present and Nd = dust grain column density (number m
-2
)
along the line of sight from object to observer. (IMA2 p. 401)

Mie Scattering compared to Rayleigh Scattering
In formulating Mie Scattering theory (Gustav Mie, 1868 - 1957), Mie assumed dust grains are spherical of
radius a and defined an extinction coefficient Q/g. His formula applies to light having a wavelength in
the same size range as the dust grain diameters (a ):
a
3
/
Thus, the Mie scattering cross section or probability for a is greater with shorter than longer wavelengths.
When viewing a nebula with scattered light, the observer often sees bluer light from scattering.
For Rayleigh scattering the scattering particles are much smaller, having a<< , and there is greater Rayleigh
scattering of shorter wavelength light. Rayleigh scattering exhibits a strong wavelength dependence, namely

-4
This causes, for instance the predominance of scattered blue light in the sky as seen from the
ground.
182

180
http://en.wikipedia.org/wiki/Extinction_%28astronomy%29
181
Bok globules:
http://www.eso.org/public/images/eso0102c/
http://iopscience.iop.org/1538-3881/124/5/2749/fulltext/202251.text.html
182
http://en.wikipedia.org/wiki/Rayleigh_scattering
Mie scattering (still with a ) is lower for longer wavelengths. Mie scattering accounts for interstellar
reddening, the reddening of starlight that occurs from passage through dust on its way to the observer. (The
amount of this effect can be differentiated from Doppler redshift by analyzing emission and absorption lines.)
For Rayleigh scattering , the lesser scattering of red light results in greater red light transmission in the sky
and redder sunsets.
In Mie scattering theory, when is large compared to a, Q goes to zero (i.e., small dust particles do not affect
long wavelength radiation). When is small compared to a, Q becomes constant and a
2
(i.e., large dust
particles simply block radiation that encounters them).

Molecular Extinction
In actual interstellar extinction measurements, there are regions or bumps of greater than expected
extinction in the UV range, which is apparently due to other molecules. The peak at 217.5 nm may be
attributable to graphite, which would therefore be a major component of ISM dust. However, it might also be
due to polycyclic aromatic hydrocarbons PAH (such as anthracene, coronene, etc.) There is also evidence of IR
extinction at 9.7m and 18 m from SiO.
Light scattered by dust tends to be slightly polarized by a few percent, and a preferred orientation suggests
the presence of a magnetic field in the ISM.
It is likely that interstellar dust partly consists of graphite and/or SiO.

Hydrogen as Dominant Component of the ISM
Hydrogen is the dominant component of the ISM, and is found in H2 (molecular), H I (neutral H) and H II
forms. It is about 70% of the ISM by mass, with He the bulk of the remainder, and much smaller fractions of
metals, carbon, silicon, etc.

Hydrogen 21 cm Emission line
The hydrogen 21 centimeter line or H I line, one of the great gifts of nature for radio astronomers, refers to the
electromagnetic radiation spectral line that is created by a change in the energy state of neutral hydrogen
atoms, H I. The photons emitted have a precise frequency of 1420.40575 MHz (a microwave-radio emission),
which is equivalent to a wavelength of 21.10611405 cm in free space. The spectral flux density of 21-cm
radiation, like other radio, microwave, and IR emissions, is often expressed in Jansky units (1 Jy = 10
26
W
m
-2
Hz
-1
, where the Hz
-1
allows inclusion of detector bandwidth).
The emission arises as a forbidden transition, therefore improbable, and occurs with a long half-life of
2.9x10
-15
s
-1
. (Thus it can be observed arising only from tenuous gases such as in the ISM, for which the
probability of atomic collisions is relatively low.) The H I atom in the ground states may be in one of two
electron spin states, that with ms = +1/2 (for which electron and proton have parallel but opposite anti-
parallel orientations in a magnetic field) and the lower energy state with ms = -1/2 (for which electron and
proton have the same and parallel orientations in a magnetic field). The flip of the spin to anti-parallel is
associated with the photon emission.
These 21-cm photons can penetrate large clouds of interstellar cosmic dust that are opaque to visible light.
The center of the 21-cm line remains optically thin over large interstellar distances, so that
Tv

where NH is the column density of H I (in atoms m
-2
) and v reflects the line broadening from Doppler effects,
typically ~10 km s
-1
... NH is generally proportional to Nd when AV is < 1,and it is inferred that dust and gas
are distributed together throughout the ISM... This relationship does not apply for denser dust (IMA2 p. 406)
Optically thick dust clouds shield hydrogen from sources of UV, allowing molecular H2 to persist. Dust can
facilitate the chemical combination of H atoms to H2 (and other molecules such as H2O) because it provides
fixed combining sites and can absorb some of the binding energy released. Molecular clouds containing H2
tend to be surrounded and shielded by H I shells.
The 21-cm emission is used in mapping the location of H I in galaxies and the ISM, in measuring radial
velocities with Doppler, and estimating magnetic fields with Zeeman effects. They are also used to assess
structure and kinematics of galaxies, including the Milky Way.

Detection of Molecular H
2

The H2 molecule does not emit the 21-cm line and is quite difficult to directly detect at the relatively cool
temperatures of the ISM. Its presence is usually inferred based on the amount of a proxy or tracer molecule
with more favorable emissions such as CO, which usually has a number abundance of 10
-4
that of H2 and an
emissions at 2.6 mm and 1.3 mm. Other useful molecular and radical tracers are OH (hydroxyl radical,
observed at 18-cm and other wavelengths), CS (carbon monosulfide), CH (methylidyne radical), C3H2
(cyclopropenylidene), HCO
+
(formyl cation), and N2H
+
(diazenylium).
It may also be helpful to study specific isotopes of certain of the atoms making up otherwise identical
molecules. The textbook refers to these tracers as isotopomers, although it appears that isotopologue would
be the preferred term.
183

Interstellar Clouds Classification
A broad and here abbreviated classification is as follows (IMA2 p. 408):
Diffuse H I Clouds: Diffuse H I clouds have temperatures of 30 to 80 K, number densities
of 1 to 8 x 10
8
m
-3
, and masses in the range of 1 to 100 M.
Diffuse Molecular Clouds (Translucent molecular clouds): These have primarily atomic H (H I) and
extinction is 1 < AV < 5. Molecular H (H2) can be observed in areas of higher column density. These are
like H I clouds but with somewhat higher masses. They have number densities of 5 to 50 x 10
8
m
-3
. and T
of 15 - 50K. The molecular gas is clumpy or filamentary on small scales, helping to explain why stars form
in clusters (ZXI lecture 2 and here
184
).
Giant molecular clouds (GMCs): These are enormous complexes of dust and gas with temperatures of ~15
K, number densities of 1 to 3 x 10
8
m
-3
, and masses typically 10
5
to up to 6 x 10
6
M, typically about 50 pc
across. They are clumpy with local regions of higher density. They may contain Dark cloud complexes (like
the Horsehead Nebula), and smaller denser clumps. The densest cores may have masses of 10 M, and
hot cores may form with masses of 10 - 3000 M and T of 100 - 300K. Based on IR observations, hot cores
probably harbor embedded massive young O and B stars. They are the breeding grounds of protostars.
Bok globules: These very dark structures typically located in H I regions outside larger molecular complexes,
have very high AV ~10, low T ~10 K, large number densities n > 10
10
m
-3
, low masses ~ 2 - 50 M and small
size < 1 pc.
185
They may be dense cores that have been stripped of their surrounding molecular gas by hot
massive stars. (IMA2 p. 409)

183
Isotopologues and Isotopomers:
Isotopologues are molecules that differ only in their isotopic composition. Simply, the isotopologue of a
chemical species has at least one atom with a different number of neutrons than the parent.
http://en.wikipedia.org/wiki/Isotopologue
Isotopomer: Isomers having the same number of each isotopic atom but differing in their positions. The
term is a contraction of 'isotopic isomer'. Isotopomers can be either constitutional isomers (e.g. CH2DCH=O
and CH3CD=O) or isotopic stereoisomers [e.g. (R)- and (S)-CH3CHDOH or (Z)- and (E)-CH3CH=CHD]
http://goldbook.iupac.org/I03352.html
184
http://en.wikipedia.org/wiki/H_II_region
185
http://en.wikipedia.org/wiki/Bok_globule

Interstellar Chemistry
Summarization omitted (seeIMA2 p. 409).

Heating and Cooling of the ISM
Heating in the ISM and ISM clouds arises from cosmic rays, namely charged particles, through the ionization
of hydrogen atoms and molecules. The ejected electrons interact with the ISM to increase kinetic energy and
T. Heating also arises from ionization of C atoms by UV from starlight, photoelectric ejection of electrons from
dust grains by UV, and ionization of H by stellar X-rays, also in some cases shocks from supernovas.
Cooling arises mostly from emission of infrared photons by atoms or molecules previously excited by
collisions, etc. These IR emissions typically can pass easily through the cloud and the ISM.

Protostars

Definition of Protostar
Protostars are pre-nuclear burning objects formed from interstellar molecular clouds (IMA2 p. 412).
According to IMA2 p. 424, the protostar stage begins when the collapsing cloud or object shrinks to a radius of
about 5 AU, light can no longer readily escape, and collapse is no longer isothermal but has become adiabatic
(so that the core is rising in temperature). See below regarding When a Protostar becomes a Star.

Timescale for Protostar Evolution
Protostellar evolution proceeds at a slower and more prolonged rate, given by the Kelvin-Helmholtz timescale
tKH (see earlier discussion, about 10
7
years for the Sun), rather than by the free-fall timescale tff (on the order
of 10
5
years, thus tKH >> tff, see IMA2 p. 416, 425).

The Jeans Criteria
Under what circumstances do protostars form from collapsing cloud globules and cores? In the theory
formulated by Sir James Jeans (1877 - 1946), he used the virial theorem and derived that if kinetic energy is
insufficient, the cloud will collapse. The condition for collapse of a cloud of mass MC is

(
T
)
/
(
)
/

where MJ is the Jeans Mass, the minimal mass necessary to initiate the spontaneous collapse of the cloud,
is the mean molecular weight, mH is the mass of a hydrogen atom = 1.67 x 10
-27
kg, and 0 is the initial
cloud density. The Jeans criterion for collapse may also be expressed in terms of the minimal cloud radius RC
for which a cloud of density 0 will collapse,
(
T
)
/

where RJ is the Jeans Length.

Bonnor-Ebert Mass
Another important factor that must be considered is an external pressure from the ISM surrounding the
cloud, such as an encompassing GMC around a dense core. With the addition of such a surrounding
pressure P0, the Bonnor-Ebert Mass criterion for collapse is
/

where
v
T/
/

and cBE 1.18. Diffuse clouds are stable against collapse, but cores of GMCs are not.

Timescale for homologous collapse
In the case of a cloud to collapse satisfying the Jeans criterion and having uniform initial density (homologous
collapse ), a calculation shows that the free-fall timescale tff for the collapse of the cloud is
)
/

In this case, all parts of the cloud take the same time to collapse and the free-fall time is independent of initial
radius. However, collapse in the case with a starting higher center density is also a possibility and will have a
shorter tff. Sample tff calculations yield 100,000 400,000 years. The rate of collapse steadily accelerates
over time.

Fragmentation of Collapsing Clouds
Clouds that are collapsing do not generally collapse into one giant protostar. Rather, the collapsing cloud
tends to fragment and preferentially form protostars in groups, a process termed fragmentation. Density of
the cloud fragments are increasing by many order of magnitude over surrounding regions. Local
inhomogeneities include regions that individually satisfy the Jeans Criterion, and these begin their own local
collapse. Probably only 1% of the collapsing cloud actually goes into star production. At some point, the
collapse is no longer isothermal and becomes adiabatic, and the rising temperature of the collapsing gas must
be partly offset by release of heat energy in order for collapse to continue. During adiabatic collapse, the
minimum value for the Jeans Mass of fragments that produce separate collapse is:
(
T
/
e
/
/
)

where e is an efficiency factor of about 0.1, and is the mean molecular weight. The interpretation is that
fragmentation ceases when the segments of the original cloud begin to reach the range of solar mass objects
[1 M]. (IMA2 p. 419) Increasing temperature halts further collapse and we have protostars. Other
sophisticated calculations yield a value of ~0.01 M as the lower limit at which fragmentation ceases.

Magnetic Fields
With collapse comes conservation of angular momentum, decreasing moment of inertia, and increasing rate of
angular rotation, introducing turbulence and probably magnetic fields. Zeeman splitting measurements show
a field of 1 to 100 nT in molecular clouds. With further collapse, the field is frozen-in, becomes stronger, and
rising magnetic pressure tends to resist further collapse.
The supercritical mass MB of a cloud, in terms of the magnetic field B, is given by
T
) (
c
)

where B is expressed in nT and R of the sphere is in pc. This is the mass above which gravitational collapse
overwhelms the resisting magnetic pressure. Below this mass, the core is subcritical.

Ambipolar Diffusion
If the magnetic field strength is locally lower, what would have otherwise been a subcritical mass may be
supercritical in that location. Neutral species (such as H I) may migrate in a preferred direction, called
ambipolar diffusion. This process can last a long time before free-fall collapse begins. (I have not fully
digested this topic.) In astrophysics, ambipolar diffusion refers specifically to the decoupling of neutral
particles from plasma in the initial stage of star formation. The neutral particles in this case are mostly
hydrogen molecules in a cloud that would undergo gravitational collapse if it were not collisionally coupled to
the plasma. The plasma is composed of ions (mostly protons) and electrons, which are tied to the interstellar
magnetic field and therefore resist collapse. In a molecular cloud where the fractional ionization is very low
(one part per million or less), neutral particles only rarely encounter charged particles, and so are not
hindered in their collapse into a star.
186

Numerical Simulation of Protostellar Evolution and the Onset of Stardom
These complex assessments are done by advanced computer models employing MHD equations . (IMA2 p.
422). Many details are omitted here, models of various types predict varying results, and I have not digested
this topic thoroughly.
Although according to IMA2 p. 424, the protostar stage begins when the collapsing cloud or object shrinks to a
radius of ~5 AU, this stage is much larger than the stages depicted in the graphs that follow
187
. These graphs
simulate the evolution of protostars of varying masses, beginning at the lower right.
When Does a Protostar Becomes a Star?>: This topic appears to be somewhat hard to pin down. Wuchterl
& Tscharnuter 2003 (their data is show in the graphs below) define stellar zero age (i.e. when a protostar
becomes a star) as follows: We propose to use as age zero the instant of time, when the interior of the gas
spheres is thermally enclosed for the first time. The enclosure means that photons cannot escape from the
entire object directly but are radiated from a photosphere. Energy transfer to the photospheric bottleneck
then delays the cooling and determines how the thermal reservoir in the interior is emptied... As a practical
definition for age zero we propose to use the instant of time when the Rosseland mean optical depth of a
gaseous object equals 2/3... For a protostar this instant of time is practically at the end of the isothermal
phase of the collapse... It corresponds to the sharp initial luminosity rise (within 1000 yr for 0.05 to 10 M
that occurs when the first (molecular hydrogen) protostellar cores form.
188
Some authors simply state the
onset of fusion reactions is the point when a protostar becomes a star. Note that the zero stellar age does not
occur at the same time as the somewhat later zero-age main sequence (ZAMS).
The following diagram depicts protostar evolution for various masses:

186
http://en.wikipedia.org/wiki/Ambipolar_diffusion
187
Wuchterl G & Tscharnuter WM, A&A 398, 1081-1090 (2003) ,
www.aanda.org/articles/aa/full/2003/06/aa8963/aa8963.right.html
188
ibid

Protostar evolution in H-R diagram for collapse and early pre-main sequence cloud fragments.
Dotted lines are isochrones. See above regarding defining zero stellar age versus ZAMS.
(Wuchterl & Tscharnuter 2003)

As protostars accrete matter, shock waves from infalling material heat the cloud and initially produce its
luminosity (i.e., fusion is not yet occurring). The core temperature steadily increases.
Eventually deuterium burning begins in the protostar (at about 10
6
K)
189
, because it has a higher reaction
cross section than the protium of the P P I chain, and deuterium is the most easily fused nucleus available in
protostars. Deuterium burning is strongly T dependent and has the effect of preventing the central T from
rising much above 10
6
K. According to Palla and Zinnecker (2002), there are four stages of D burning:
190

1) steady state central (core) D burning, fed by infall from convective eddies,
2) onset of a radiative barrier around the central region, preventing further infall of matter;
3) interior region depleted of deuterium, central deuterium burning stops, allowing the central
temperature of the protostar to (?) increase, and
4) shell burning of deuterium in expanding convective shells outside the D-depleted central region and
radiative barrier.
When central deuterium burning ends, a temporary dip in luminosity and T results, and the star swells
temporarily delaying gravitational contraction. After this, T rises, the convecting core becomes radiative, and
more steady burning by P P I chain begins.
191

The following graph (annotated by MCM from Palla 1999)
192
roughs out the sequence in an accreting protostar
from onset of deuterium burning to onset of ZAMS of a main-sequence star. Palla states, The two curves
refer to the evolution computed assuming that matter is accreted directly onto the protostar through a strong
shock and to the case where the core surface is treated as a normal photosphereas in the case of accretion
through a circumstellar disk.

189
http://en.wikipedia.org/wiki/Deuterium_burning
190
Palla F; Zinnecker H (2002). Physics of Star Formation in Galaxies. Springer-Verlag. pp. 2122
191
Ibid.
192
Palla F, The Evolution of Pre-Main-Sequence Stars, 1999.
www.cfa.harvard.edu/events/1999/crete/palla_CreteII.ps
Here, a protostar, starting at a much smaller
radius than it had at its inception (i.e., after
considerable contraction), accretes from 0 M
at the start to 8 M. The open circles and
intervals between them show the onset of
central deuterium burning, the expansion of
the protostar during this burning, onset of
Deuterium shell burning with further rapid
expansion, depletion of deuterium with
contraction, onset of protium
1
H burning first
with the CNO cycle and eventually with the PP
cycle leading at last to the onset of the Zero
Age Main Sequence point. See citation above.
The observational search for protostars is
underway. They may have a characteristic
spectral line profile (IMA2 p. 424): blueshift
from infall on the far side, redshift from infall
on the near side, and a centrally positioned
absorption dip arising from the cloud located
between the observer and the center but far
out from the region of central collapse.

Pre-Main-Sequence Evolution, Young Stellar Objects, and the ZAMS

Not all models give the same results in this rapidly evolving field.
The zero-age main sequence ZAMS is the point in time when a star reaches a state of stable equilibrium in
hydrogen (
1
H) burning for which this is the dominant process of energy production, rather than release of
gravitational energy, and for which there is hydrostatic equilibrium (IMA2 p. 429 and here.
193
) On the H-R
diagram, the locus of ZAMS for stars of varying masses is a roughly diagonal line passing from lower right to
upper left.
The tracks in the graph to the right (Palla
and Stahler 1993) depict ZAMS positions
on the H-R diagram, namely at the left end
of each track (see the caption).
194
Tracks
begin at the birthline (complex, see article)
of the protostar. Most normal main
sequence stars fall slightly up and to the
right of their ZAMS position.
The times for stars to reach ZAMS vary
inversely with mass: Time to ZAMS is
28,000 yr for a 60 M star (virtually no
PMS phase at all), 40 Ma for a 1 M star ,
and 68 Ma for a 0.8 M star (starting at
the Hayashi Track, IMA2 p. 427).
There is a period of time for which a
protostar has become a star but is not yet
at the ZAMS. This interim stage is termed
a Pre-main-sequence star (PMS).
A Pre-main-sequence star (PMS) is a star
that has not yet reached the main
sequence. It can be a T Tauri star (young,
< 2 M), an FU Orionis star (< 2 M), or a
Herbig Ae/Be star (2 to 8 M). The energy
source of these objects is primarily
gravitational contraction (as opposed to
hydrogen burning in main-sequence
stars)... PMS stars can be differentiated
from main-sequence dwarf stars by using
stellar spectra to measure the correlation
between gravity and temperature. A PMS
star will have a larger radius than a main-
sequence star, and thus be less dense...
Pre-main-sequence stage will last less than
1% of a star's life (in contrast, the star will
spend about 80% of its life on the main sequence).
195

Classical calculations showed that a PMS comes to have its outer atmosphere dominated by H
-
ion which
adds opacity and leads to a convective outer atmosphere. C. Hayashi showed that PMS evolution in this state
along the Hayashi Track is almost vertical on the H-R diagram (luminosity is decreasing while Teff is only
minimally rising).
In the following H-R graphs
196
annotated by MCM, the Hayashi Tracks appears as downstrokes on the far
right of each curve. At the lowest point of the Hayashi track, the radiative core allows energy to escape into

193
http://en.wikipedia.org/wiki/Zero_age_main_sequence
194
Palla F; Stahler SW, The Pre-Main-Sequence Evolution of Intermediate-Mass Stars, Astrophysical Journal
v.418, p.414, 11/1993, http://adsabs.harvard.edu/doi/10.1086/173402
195
http://en.wikipedia.org/wiki/Pre-main-sequence_star
the convective envelope, leading to rising luminosity. At about this minimum point,
1
H nuclear reactions
begin in earnest, namely PP I and CNO cycle, while gravitational collapse is providing less of a contribution to
L. At the left-hand peak of the luminosity (marked by the short dotted line where the convective core first
appears), the central core begins to expand, luminosity and teff start to decrease. The star thereby approaches
the ZAMS state of stable
1
H burning (not explicitly marked but apparently at approximately the left-sided
terminus of the tracks).

Optimal calculations should include effects of rotation, magnetic fields,
mass loss or accretion, external pressure, feedback caused by infalling
matter. Birth lines may be needed.
For lower mass stars (0.072 0.5 M), the curves are slightly different in
that they lack the upstroke just prior to the ZAMS (due to a lack of
efficient
12
C burning in the core).
The graph to the right is a much simplified depiction of pre-ZAMS
evolutionary tracks for stars near 1 M.
197

196
Bernasconi PA; Maeder A, About the absence of a proper zero age main sequence for massive stars,
Astronomy and Astrophysics, v.307, p.829-839, 1996.
197
www.castlerock.wednet.edu/HS/stello/Astronomy/TEXT/CHAISSON/BG311/HTML/BG31104.HTM
Massive Stars
For very massive stars, the central temperature quickly becomes high enough to burn
12
C and to convert
1
H
to
3
He. , using the full CNO cycle. The core remains convective even after ZAMS. The current largest star
known is the blue hypergiant R136a1, weighing in at 265 M and with L = 8.7x10
6
L. It is a WolfRayet star
with a surface temperature over 50,000 K.
198

It is possible that stars > 10 M do not form by accretion but by mergers of smaller stars in dense protostellar
environments (IMA2 p. 429), but the presence of an accretion disk can reduce the need for mergers.

Brown Dwarfs
Stars with mass < 0.072 M have some nuclear burning, but insufficient to be a main sequence star. Lithium
burns at above 0.06 M and deuterium burns above 0.013 M but
1
H protium burning is not sustained at
mass < 0.072 M. For objects 0.013 M < M < 0.072 M, these are termed brown dwarfs. Here, 0.013 M is
about 10 times the mass of Jupiter which is 9.54 x 10
-3
M. These objects are cool, with spectral types L
and T.
199
They are probably common, but are hard to detect due to low luminosity.

Initial Mass Function IMF
The IMF is the probability distribution of mass as a function of stellar mass m [not magnitude], given by
dN = (log m) d(log m) = k1M
-p

where dN is the number of stars found in logarithmic mass range from log m to log m + d(log m). It is plotted
typically in log-log axes. Massive stars are numerically quite rare, while small stars are abundant. A simple
power law is often assumed, such as the exponent p = 1.35 utilized by E. E. Salpeter.
200

Schematic of idealized IMF (from a full review of IMF, Bastian et al 2010
201
)

198
http://en.wikipedia.org/wiki/R136a1
199
http://en.wikipedia.org/wiki/Brown_dwarf
200
http://astro1.physics.utoledo.edu/~megeath/ph6820/lecture12_eqn.pdf
The derived present-day mass function of a sample of young star-forming regions ..., open
clusters spanning a large age range ..., and old globular clusters... Additionally, we show the
inferred field star initial mass function (IMF)... The gray dashed lines represent tapered power-
law fits to the data... The black arrows show the characteristic mass of each fit (mp), the
dotted line indicates the mean characteristic mass of the clusters in each panel, and the
shaded region shows the standard deviation of the characteristic masses in that panel... The
observations are consistent with a single underlying IMF, although the scatter at and below the
stellar/substellar boundary clearly calls for further study. The shift of the globular clusters
characteristic mass to higher masses is expected from considerations of dynamical evolution.
(also from Bastian et al 2010)

The distribution of stellar masses that form in one star formation event in a given volume of space is called
the initial mass function (IMF). The IMF has been estimated from low-mass brown dwarfs to very massive
stars. Combining IMF estimates for different populations in which the stars can be observed individually
unveils an extraordinary uniformity of the IMF. This general insight appears to hold for populations including
present-day star formation in small molecular clouds, rich and dense massive star-clusters forming in giant
clouds, through to ancient and metal-poor exotic stellar populations that may be dominated by dark matter.
This apparent universality of the IMF is a challenge for star formation theory, because elementary
considerations suggest that the IMF ought to systematically vary with star-forming conditions.
202

201
Bastian N, et al, A Universal Stellar Initial Mass Function? A Critical Look at Variations, Annual Review of
Astronomy and Astrophysics Vol. 48: 339-389, 2010, DOI: 10.1146/annurev-astro-082708-101642.
202
Kroupa P, The Initial Mass Function of Stars: Evidence for Uniformity in Variable Systems, Science 4
January 2002 Vol. 295 no. 5552 pp. 82-91 DOI: 10.1126/science.1067524
H II Regions (Ionized Hydrogen)
Hot O and B stars emit such energetic UV (> 13.6 eV) that surrounding neutral H I gas is ionized to H II.
Some of the ions undergo recombinations and de-excitations associated with emission of visible light, in
particular the Balmer series H, causing a beautiful red fluorescence. Examples include the Orion Nebula
M42. These regions are called emission nebulae, some of which are called planetary nebulae. The size of an H
II region is estimated by the Strmgren radius (IMA2 p. 432), which can yield values from 0.1 pc to 100 pc.

H II Region in Orion Nebula M42 NGC 1976
associated with the Trapezium cluster and a Giant Molecular Cloud.
This nebula in Orions sword, south of Orions belt.
Distance 1500 ly; width 30 arcminutes (4.0 parsecs) square.
Image by NASA/HST, see footnote, 2006 mosaic composite of 5 wavelengths
203
.

203
Orion Nebula:
Specific H II Region view shown above: http://www.spacetelescope.org/images/heic0601a/
Other views of Orion Nebula: http://www.spacetelescope.org/images/?search=orion+nebula
The Orion Nebula is a picture book of star formation, from the massive,
young stars that are shaping the nebula to the pillars of dense gas that may
be the homes of budding stars. The bright central region is the home of the
four heftiest stars in the nebula. The stars are called the Trapezium
because they are arranged in a trapezoid pattern. Ultraviolet light
unleashed by these stars is carving a cavity in the nebula and disrupting
the growth of hundreds of smaller stars. Located near the Trapezium stars
are stars still young enough to have disks of material encircling them.
These disks are called protoplanetary disks or proplyds and are too small to
see clearly in this image. The disks are the building blocks of solar
systems... The bright glow at upper left is from M43, a small region being
shaped by a massive, young star's ultraviolet light. Astronomers call the
region a miniature Orion Nebula because only one star is sculpting the
landscape. The Orion Nebula has four such stars. Next to M43 are dense,
dark pillars of dust and gas that point toward the Trapezium. These pillars
are resisting erosion from the Trapezium's intense ultraviolet light. The
glowing region on the right reveals arcs and bubbles formed when stellar
winds - streams of charged particles ejected from the Trapezium stars -
collide with material.
204

Large stars, especially O and B star groups called OB associations, cause
nearby dust clouds to vaporize, molecules to dissociate, and eventually
atoms to ionize, creating H II regions from H I regions. High radiation
pressure disperses much of the cloud, halting further protostar formation,
and possibly even causing loosely bound clusters to become unbound. This
is observed in Carina Nebula from nearby star Eta Carinae, and the Eagle
Nebula M16 from several massive stars:
Stellar Spire in Eagle Nebula (upper right image): A torrent of ultraviolet
light from a band of massive, hot, young stars [off the upper
edge of this image] is eroding the pillar of gas and dust (like an
eroded desert hoodoo). Distance ~6500 ly, width 6 ly, column
height 9.5 ly. (HST composite image)
205

T Tauri Stars
Graphs to right show Evolutionary Tracks for T Tauri stars.
Short dashes line = ZAMS.
206

These are low-mass pre-main sequence stars (or objects or
protostars) representing a transition from stars shrouded in
dust (visible only in IR) to main sequence stars. Their masses
are 0.5 to about 2 M. Many emit emission lines including H,
Ca II (H and K), and iron lines, along with lithium absorption.
Many emit forbidden lines [O I] and [S I], suggesting very low
density.
Inspection of the positions of T Tauri stars in the H-R diagram
with stellar evolution models (see graphs above) places them in
the region of the Hayashi Track or the trough at the bottom
ends of the Hayashi Tracks.
207
The ZAMS are found along the
dotted line lower left, and the circles represent classical and

204
Text regarding specific view shown above: http://www.spacetelescope.org/images/heic0601a/
205
http://www.spacetelescope.org/images/heic0506b/ 25 April 2005
206
Bertout, Claude, T Tauri stars - Wild as dust, Annual review of astronomy and astrophysics. Volume 27
1989, p. 351-395
207
ibid.
weak-line T Tauri stars. In other words, T Tauri stars arise and are seen before the star evolves to the ZAMS
stage.
The H lines of T Tauri stars often exhibit a P Cygni profile, suggesting an expanding shell about the object,
consisting of:
a blueshifted absorption trough due to approaching H
on the near side of shell, lying directly on the line of
sight in front of the star, and which therefore absorbs
some of the stars light

a broad unshifted emission peak from edges of shell
having with no radial movement, and

a redshifted emission from the part of the receding
rear shell that is not obscured by the star.
Their spectra can rapidly change, suggesting these
objects are very unstable.

Image to right depicts a P Cygni Profile.
208

T Tauri stars are premain sequence stars the youngest visible F, G, K, M spectral type stars (<2 Solar
mass). Their surface temperatures are similar to those of main sequence stars of the same mass, but they are
significantly more luminous because their radii are larger. Their central temperatures are too low for
hydrogen fusion. Instead, they are powered by gravitational energy released as the stars contract towards the
main sequence, which they reach after about 100 million years. They typically rotate with a period between
one and twelve days, compared to a month for the Sun, and are very active and variable.
209

FU Orionis Stars
These appear to be T Tauri stars that undergo rapid accretion of mass and marked increase in luminosity,
perhaps due to instabilities in a T Tauri accretion disk. (IMA2 p. 437) They have mass < 2 M.

Herbig Ae/Be star
A Herbig Ae/Be star (HABe) is a premain sequence star of mass 2 to 10 M, often enveloped in gas and dust.
They are a young (<10Myr) star of spectral types A or B. These stars are still embedded in gas-dust envelopes
and may be surrounded by circumstellar disks. Hydrogen and Calcium emission lines are observed in their
spectra. They are 2-8 Solar mass objects, still existing in the star formation (gravitational contraction) stage
and approaching the main sequence (i.e. they are not burning hydrogen in their center). In the Hertzsprung
Russell diagram these stars are located to the right of the main sequence. They are named after the American
astronomer George Herbig (b. 1920), who first distinguished them from other stars in 1960... Sometimes
Herbig Ae/Be stars show significant variability.
210

HerbigHaro objects (HH)
These are jets ejected in opposite directions that may be associated with pre-main sequence evolution and
young protostars, such as T Tauri stars. The rapidly moving jets excite emission lines on collisions with gas
etc. (such as seen with HH47). The jets may be associated with many of the characteristics of protostars,
including emission lines, mass loss, jets, and possibly luminosity variations (IMA2 p. 437).

208
http://en.wikipedia.org/wiki/P_Cygni
209
http://en.wikipedia.org/wiki/T_tauri_star
210
http://en.wikipedia.org/wiki/Herbig_Ae/Be_star
HerbigHaro objects (HH) are small patches of nebulosity associated with newly born stars, and are formed
when gas ejected by young stars collides with clouds of gas and dust nearby at speeds of several hundred
kilometers per second. HerbigHaro objects are ubiquitous in star-forming regions, and several are often seen
around a single star, aligned along its rotational axis... HH objects are transient phenomena, lasting not
more than a few thousand years. They can evolve visibly over quite short timescales as they move rapidly
away from their parent star into the gas clouds in interstellar space (the interstellar medium or ISM). Hubble
Space Telescope observations reveal complex evolution of HH objects over a few years, as parts of them fade
while others brighten as they collide with clumpy material in the interstellar medium... The first astronomers
to study them in detail were George Herbig and Guillermo Haro, after whom they have been named.
211

Young Stars with Circumstellar Disks
Young stars often have these disks, such as Vega
212
and -Pictoris
213
and HH30
214
. The disks may be accretion
disks, and clumps of material fall from the disk into the star. In some cases they may be debris disks (i.e.,
disks formed by fragmentation of previously formed objects by collisions). Protoplanets may be present. The
central star may be masked by dust (as is the case with HH30).
Circumstellar disks appear to be common during collapse of protostars. This is predicted by the spin-up of
the cloud during collapse to conserve angular momentum. The typically slow rotation of main sequence stars
might seem inconsistent, but suggests that angular momentum has been transferred away from collapsing
protostars and stars that would otherwise be rotating rapidly. This may take place with coupling of magnetic
fields in the stars convection zones to the ionized stellar winds, or other mechanisms. (IMA2 p. 441 and
here
215
)

Proplyds
These are probably circumstellar disks and appear to be protoplanetary disks associated with young stars of
age < 1 million years. The disks have mass much greater than 4 M.
216

211
http://en.wikipedia.org/wiki/Herbig%E2%80%93Haro_object
212
http://www.solstation.com/stars/vega.htm
213
http://www.solstation.com/stars2/beta-pic.htm
214
Watson AM and Stapelfeldt KR, Asymmetry and Variability in the HH 30 Circumstellar Disk
The Astronomical Journal Jan 22, 2007 p. 133 845, http://iopscience.iop.org/1538-3881/133/3/845
215
Lignieres F, Catala C, & Mangeney A., Angular momentum transfer in pre-main-sequence stars of
intermediate mass., Astronomy and Astrophysics, v.314, p.465-476, 1996
http://articles.adsabs.harvard.edu/full/1996A%26A...314..465L
216
A Megagerie of 30 Proplyds:
http://www.spacetelescope.org/static/archives/images/screen/heic0917aa.jpg
Main Sequence and Post-Main-Sequence Stellar Evolution (Chapter 13)

Main sequence stars (MSS) are characterized by their burning of hydrogen to helium in their cores. A star
leaves the main sequence when its core is no longer burning hydrogen (IMA2 p. 457). Here we follow these
stars, beginning at the zero-age main sequence (ZAMS) previously defined. This is a very actively studied and
rapidly evolving field for which various complex models are not necessarily in agreement. I cant pretend to
have the latest or definitive information here.

Evolution of the Main Sequence

Stellar Evolution Timescales
For pre-main sequence (PMS) stars, their timescales are the free-fall timescale tff and the thermal Kelvin-
Helmholtz timescale (tKH, about 10
7
yr for the Sun).
For main sequence stars (MSS), their timescales are also governed by the durations of thermonuclear
reactions (about 10
10
yr for the Sun). This much longer timescale accounts for why 80% of stars in our
neighborhood are MSSthis stage simply lasts longer.

Width of the Main Sequence
The main sequence is properly depicted in H-R diagrams as a broad band (see A modern Hertzsprung
Russell diagram above and here
217
) because the actual distribution of observed stars is broad. The variation
is due to differences in chemical composition, stage of evolution, and observational imprecision, etc.

Graphs of Overall Stellar Evolution
The following images are modified from figs. 13.4 and 13.5, IMA2 p. 458-9, but obtained from another source,
annotated by MCM:
218

217
http://msedblogwaterfilter.blogspot.com/2011/01/review-questions-chapters-15-16-nature.html
218
Images are modified from figs. 13.4 and 13.5, IMA2 p. 458-9, but obtained from
www.physics.sfsu.edu/~fischer/courses/Astr420/lectures/Ch13.pdf

Evolution of low-mass Sun-like 1 M stars from ZAMS to the final White Dwarf stage
ZAMS = zero-age main sequence (ZAMS);
SGB = SubGiant Branch;
RGB = Red Giant Branch;
E-AGB = Early Asymptotic Giant Branch;
TP-AGB = Thermal Pulse Asymptotic Giant Branch;
Post-AGB = Post-Asymptotic Giant Branch;
PN = Planetary Nebula
(Temp increases toward the left; image annotations in red added by MCM)
219

219
ibid.

Evolution of intermediate-mass 5 M stars from ZAMS to the final White Dwarf stage
ZAMS = zero-age main sequence (ZAMS); SGB = SubGiant Branch;
RGB = Red Giant Branch;
HB = Horizontal Branch (blueward and redward portions);
E-AGB = Early Asymptotic Giant Branch;
TP-AGB = Thermal Pulse Asymptotic Giant Branch;
Post-AGB = Post-Asymptotic Giant Branch;
PN = Planetary Nebula
(Temp increases toward the left; image annotations in red added by MCM)
220

Low-Mass Main-Sequence Evolution (including of the Sun) from ZAMS to the End of MSS
These are 0.3 M to 1.2 M. In the upper part of this mass range, stars including the Sun have a radiative
core because for much of their time as MSS, they are not emphasizing the temp-dependent CNO cycle (as is
the case for stars > 1.2 M. The lower mass stars in this range have convective cores.
Rising T, R, and L: Recall that L and R have steadily risen since ZAMS, at least for the Sun (IMA2 p. 350).
The Sun is thought to be 4.57 Ga from ZAMS. The rise in L and R occurs because the fission reactions

220
ibid.
increase mean molecular weight in the core (IMA2 p. 293). According to the ideal gas law previously
presented (IMA2 p. 291), if density and T were held constant, and rises, contraction in the core occurs and
gravitational energy is released, increasing interior T. Rising T increases the radius inside of which nuclear
reactions can occur. The rate of the PP reaction also increases with rising T. The increased energy output
propagates to the surface, and thus the stars luminosity and Teff gradually increase.
The evolutionary tracks of MS stars from 0.8 to 25 M are nicely depicted in detail in fig. 13.1 of IMA2 p. 448,
starting at the ZAMS = point 1. The table 13.1 accompanying fig. 13.1 (IMA2 p.449) show dramatic
differences in the burning time for larger vs. smaller stars. The Sun at 1 M reaches point 2 in 7048 Ma
whereas the 0.8 M star reaches point 2 in 18829 Ma and a 25 M star reaches point 2 in only 6.3 Ma. (The
criteria for point 2 is not stated, but for larger stars it appears to be the point where decline of Teff ceases and
Teff begins to rise.) Our Sun is currently on the path between 1 and 2. The interior physical and isotopic
structures of the present-day Sun (parameters graphed as a function of radius) are depicted in IMA2 p.450
(based on Bahcall 2001), and of the Sun near point 3 (about 9.8 Ga from ZAMS) in IMA2 p. 451 (based on
Iben 1967). Comparing these graphs, central depletion of
1
H becomes apparent.
By point 3, nuclear burning continues in a shell surrounding the
1
H depleted core (from which no further
luminosity is arising), but overall and surface luminosity continues to rise. The star moves further to the red
in the H-R diagram (Teff decreases) as R and L increase, but this is still a MSS.
Mario Schnberg (19141990) and Subrahmanyan Chandrasekhar (19101995) calculated in 1942 the
SchnbergChandrasekhar limit (SC limit). This is the maximum fraction of a stars mass, typically 10 to
15%, that can exist as an isothermal burnt out helium core (too cold for triple alpha burning), and for which
the core is still able to support the enveloping material above it. If this limit is exceeded, as can only happen
in massive stars, the core collapses, releasing energy that causes the outer layers of the star to expand to
become a red giant.
221
(Note: this limit is not the same as the Chandrasekhar limit, which is the maximum
mass of a stable white dwarf star, about 1.4 M, beyond which, a neutron star or black hole may form.)
The calculation for SC limit is based on the mean molecular weights of the envelope versus the isothermal
core (IMA2 p. 451):
(

The core increases in mass as the shell burns its way outwards through the star. If the star's mass is less
than approximately 1.5 solar masses, the core will become degenerate before the SchnbergChandrasekhar
limit is reached. But if the mass is greater than approximately 6 solar masses, gravitational collapse [on a
Kelvin-Helmholtz timescale] will release so much energy that the core will not become isothermal prior to the
start of helium fusion. In the remaining case, where the mass is between 1.5 and 6 solar masses, the core
will grow until the limit is reached, at which point it will contract rapidly until helium starts to fuse in the
core.
222

When this SC limit is reached (point 4 in the diagram), a star of < 1.2 M is no longer a MSS (IMA2 p. 452)
and continues to redden with falling Teff. The core acquires partial degeneracy, the percent being greater in
less massive stars. Degeneracy is a state wherein the electrons occupy the lowest possible energy levels,
allowing for the Pauli exclusion principle for fermions, and resist being packed in any tighter. Degeneracy
provides Temp-independent pressure resisting collapse, and smaller stars may never reach the SC limit before
the next stage of nuclear burning begins, because of degeneracy pressure.

Main-Sequence Evolution of Massive Mass Stars from ZAMS to the End of Main Sequence Star
These Massive Mass Stars are here defined as > 1.2 M. Unlike those just below 1.2 M, these have a
convective core which keeps their material well mixed. When a 5 M star reaches point 2 (having X, = 0.05,
the mass fraction of
1
H), the entire star begins to contract (presumably the definition of point 2). L rises from
gravitational energy release, and Teff rises. The graph thus bends up and to the left. For a star of > 1.2 M,
this point 2 is the end of the MSS phase.

221
http://www.daviddarling.info/encyclopedia/S/Schonberg-Chandrasekhar_limit.html
222
http://en.wikipedia.org/wiki/Sch%C3%B6nberg%E2%80%93Chandrasekhar_limit
Late Stages of Stellar Evolution (Post-Main Sequence)

A star leaves the main sequence when its core is no longer burning hydrogen and consists mainly of He (IMA2
p. 457). The textbook considers stars of 1 and 5 M mass for detailed evolution descriptions, which I have
simplified somewhat.
For a 1 M Post-MS star, the core begins to contract, H-burning takes place in shells outside the core,
luminosity increases, R increases slightly, and Teff decreases. The star becomes a red giant and evolves up
and to the right on the H-R diagram.
For a 5 M star, the entire star begins to contract, releasing gravitational potential energy. The L rises
slightly, R decreases, and Teff rises. Eventually, a shell of H starts to burn, the outer envelope expands, L
decreases, Teff drops.
For both 1 M and 5 M stars, the helium core increases in mass from hydrogen shell burning. The envelope
expands, Teff drops, and redward evolution occurs across the Subgiant Branch SGB. For the 5 M star, there
is actually a terminal temporary drop in L (point 5 in the figure 13.1 graph).
The expansion of the stellar envelope and the decline in Teff causes a rise in photospheric opacity from H
ion,
causing a convection zone to form near the surface for both sizes, a zone extending deep into the star. The
star begins to rise rapidly in L with modest further drop in Teff along the Red Giant Branch RGB (a path
similar to the Hayashi track of pre-MS stars).
Deepening convection mixes previously unmixed regions, and lithium burning occurs and Li becomes widely
depleted. The photospheric abundance of Li decreases, and
3
He increases. Several other isotopic abundance
ratios change (IMA2 p. 461). The observable transport of materials by convection from the deep interior to the
surface is termed the First Dredge-up Stage. The RGB is a time of hydrogen-burning shell.
Luminosity rises until the Red Giant Tip of the RGB phase is reached.
In stars of < 1.8 M but greater than about 0.5 M, the release of triple alpha fusion energy in the core is
delayed by core cooling. The core becomes strongly electron degenerate. The helium [in the core] is
degenerate, meaning it is supported against gravity by quantum mechanical pressure rather than thermal
pressure. Thus an increase in the temperature in the material undergoing fusion does not act to expand the
material and by doing so cool, and there is no regulation of the rate of fusion.
223
Eventually the triple alpha
fusion reaction begins in the degenerate core, and is very transiently explosive (termed the Helium Core
Flash, in which the core luminosity is extremely high, up to 10
11
L, but lasting for only a few seconds).
Much of the energy released is absorbed in the outer layers. This discontinuity makes further modeling for
such low mass stars difficult until quiescent He core burning and H shell burning conditions arise.
In larger stars such as for the 5 M star illustrated, the T (1.3 x 10
8
K) and density are high enough for QM
tunneling to overcome the
4
He nuclear Coulomb barrier, allowing triple alpha He burning fusion processes to
begin (generating
12
C, some of which combine with alphas to yield
16
O). A star with mass greater than about
2.25 solar masses starts to burn helium without its core becoming degenerate and so does not exhibit this
type of helium flash.
In very low mass stars (< ~0.5 M), the core is never hot enough to ignite helium burning. The degenerate
helium core will keep on contracting, and finally becomes a helium white dwarf.
224

During He burning, the core and the hydrogen shell expand, cooling it and abruptly decreasing L. The
envelope contracts, and Teff rises, and the star moves along the lower blueward limb of the Horizontal
Branch HB (which occurs for both low and intermediate mass stars). This phase is the helium-burning
analog of the hydrogen-burning main sequence, but lasts for a much shorter time. The furthest point to the
left (hottest Teff) on the HB is attained: before the He has been depleted; the mean molecular weight has
increased; the core is contracting; and the envelope is cooling and expanding. The upper and redward limb
of the HB then begins, and soon the core He is depleted. Many stars exhibit instabilities during this phase,
exhibiting pulsations in L, R, T, etc.
Asymptotic Giant Branch AGB: When luminosity begins to rise, the Early Asymptotic Giant Branch EAGB
has been reached. (The AGB is so named for seeming to approach asymptotically to the RGB.) Energy output

223
http://en.wikipedia.org/wiki/Helium_flash
224
Example with M = 0.18-0.19M: http://arxiv.org/abs/astro-ph/0404291
is dominated by the He burning shell just outside the core, the core is mainly C and O with little He
remaining, the envelope is expanding, L is rising, Teff is declining somewhat (initially). The convective envelope
deepens, a Second Dredge-up of interior isotopes occurs, and this brings He and N to the envelope.
The Thermal-Pulse Asymptotic Giant Branch TP-AGB ensues, with H-shell burning reignited. The more
deeply positioned He-burning shell exhibits quasiperiodic On/Off burning, termed Helium Shell Flashes,
which are associated with dumping of helium ash to the lower layers and abrupt rises in L... Pulses last a
few hundred years.
225
The period between pulses ranges from thousands of years for stars of 5 M to
hundreds of thousands of years for low mass stars of 0.6 M. Iben has given us a model (see graph to follow)
showing for a 0.6 M star over 2.4 million years the predicted helium flashes, occurring in similar but not
identical intervals of about 130,000 - 325,000 years. A similar model for a larger star is depicted in the graph
that follows the first graph:

This graph is extracted from an article by Iben 1982
226
regarding modeling of Helium Shell Flashes
on the Thermal-Pulse Asymptotic Giant Branch of a 0.6 M star. I have simplified it
by deletion of the upper 1/2 of the graph that depicts helium flashes from 0 to 11.5 x10
5
yr.

225
http://en.wikipedia.org/wiki/Helium_flash
226
Iben, I., Jr., Low mass asymptotic giant branch evolution. I, ApJ, Part 1, vol. 260, 1982, p. 821-837.

This graph is extracted from Ibens 1975 article
227
regarding modeling of Helium Shell Flashes
(relaxation oscillations) on the Thermal-Pulse Asymptotic Giant Branch of a 7 M star.
I have simplified it by deleting information on parameters other than luminosity and Teff:

IMA2 p. 465 raises the possibility that certain pulsating Long Period Variables (LPVs), which have periods of
100 to 700 days, including the subclass of Mira variable stars, may be experiencing shell flashes that could
account for the observable changes in periods, etc.
A Third Dredge-up may occur in conjunction with the helium flashes. For stars > 2 M, the convection zones
will extend deep enough to dredge up Carbon. This can cause the spectrum of an oxygen-rich giant star (for
which NO > NC) to transform into that of a carbon star (for which NC > NO). Such carbon stars are designated
with spectral class C. There are also spectral class S stars for which ZrO lines are present and in some cases
replacing TiO lines.
S-Process Nucleosynthesis: Some TP-AGP stars exhibit
99
Tc in their spectra, a radioisotope with half-life of
only 211,000 years, indicating it must have relatively recently been dredged up from the deeper sites of
nucleosynthesis. (This is not the metastable form of technetium
99m
Tc which has a much shorter half-
life and is used for medical diagnosis.)
99
Tc is formed by s-Process nucleosynthesis (here, s=slow). In this
slow process, a slow neutron is captured by the nucleus of a stable isotope that then decays by beta-decay to
a stable daughter having 1 greater atomic number than the parent. These decays to new isotopes are able to
occur when the density of slow neutrons is sufficiently low, allowing the decay to occur before another
neutron can be captured.
This s-process accounts for about 1/2 of the isotopes synthesized [in stars] that are heavier than iron. The
s-process is believed to occur mostly in Asymptotic Giant Branch stars. In contrast to the r-process [r=rapid]
which is believed to occur over time scales of seconds in explosive environments [such as core-collapse
supernovae Types Ib, Ic and II], the S-process is believed to occur over time scales of thousands of years. The
extent to which the s-process moves up the elements in the chart of isotopes to higher mass numbers is
essentially determined by the degree to which the star in question is able to produce neutrons, and by the
amount of iron in the star's initial abundance distribution. Iron is the "starting material" (or seed) for this
neutron capture - beta-minus decay sequence of synthesizing new elements.
228
Neutron arise in stellar
nuclear reactions primarily by the following reactions,
+ e
+
and
e + e
+

227
Iben, I., Jr., Thermal pulses; p-capture, alpha-capture, s-process nucleosynthesis; and convective mixing
in a star of intermediate mass, Astrophysical Journal, vol. 196, Mar. 1, 1975, pt. 1, p. 525-547., 3/1975.
228
http://en.wikipedia.org/wiki/S-process
(IMA2 p. 313 shows 2 other stellar reactions that yield neutrons.) Because of low neutron fluxes, the s-
process is not capable of producing any significant amounts of heavy radioactive isotopes such as thorium or
uranium.
229
It terminates at the heaviest stable isotope it can generate, namely
209
Bi, although the
culminating s-process cycle involves
210
Bi,
210
Po,
206
Pb, and
209
Pb.
Mass Loss and AGB Evolution: Stars in the AGB phase lose mass rapidly by expelling large amounts of
matter including dusts of silicates or carbon, helping to enrich the ISM. The possible causes of mass loss
include high stellar wind, helium shell flashes, and pulsations of the envelope in LPVs. In addition, as the
stars expand, the surface gravity diminishes, lowering the binding energy.
Stars of ZAMS mass > 8 M will eventually supernova, and will be discussed later. Stars of ZAMS mass < 8
M continue to synthesize C and O in the He-burning shell. In the core, increasing density leads to rising
and dominance of pressure by electron degeneracy. For stars of ZAMS mass < 4 M , the C and O in the core
will never burn. For stars of ZAMS mass 4 to 8 M, if there had been no mass loss, the core would reach a
point where it could not maintain hydrostatic equilibrium, despite the additional electron degeneracy
pressure, and a catastrophic core collapse would occur. Instead however, the high mass loss rate has
reduced mass sufficiently to prevents collapse. The core synthesizes more oxygen, neon, and magnesium
(becoming a ONeMg core) and stellar masses at this stage are below the Chandrasekhar limit of 1.4 M.
The intense superwind of around 10
-4
M yr
-1
may be responsible for OH/IR sources. OH radicals (which
are unstable molecules) are detected by infrared emissions of maser transitions from a long-lived metastable
state to ground state. (Galactic superwinds with OH megamasers have also been described.
230
)
Post-Asymptotic Giant Branch (post-AGB): Eventually, the cloud surrounding the supergiant star thins and
a supergiant star typically of F or G spectral class becomes apparent. The track for supergiants turns
blueward, moving nearly horizontally (constant luminosity). The remainder of the stars envelope is eventually
expelled, the H and He burning shells are exhausted, and the star drops rapidly in luminosity until it
becomes a white dwarf star (WD). A white dwarf has a degenerate C-O core surrounded by a thick layer of
residual H and He. These are the final remnants of ZAMS stars of M < 8 M . For instance, a star of 0.6 M
at ZAMS becomes a WD with radius 0.0285 R.
Planetary Nebula: A glowing ionized expanding shell of gas may be seen around a WD progenitor, termed a
planetary nebula (a misnomer left over from before 1785). These are often multi-colored due to emissions in
visible light of multiply ionized species. For example, blue-green arises from 500.68 nm and 495.89 nm
forbidden lines of [O III], etc.
Planetary nebulae play a crucial role in the chemical evolution of the galaxy, returning material to the
interstellar medium that has been enriched in heavy elements and other products of nucleosynthesis (such as
carbon, nitrogen, oxygen and calcium). In more distant galaxies, planetary nebulae may be the only objects
that can be resolved to yield useful information about chemical abundances.
231

An image of a prominent planetery nebula follows.

229
ibid.
230
Baan WA, et al, Molecular outflows in powerful OH megamasers, Astrophysical Journal, Part 1 vol. 346,
Nov. 15, 1989, p. 680-689, http://adsabs.harvard.edu/abs/1989ApJ...346..680B
231
http://en.wikipedia.org/wiki/Planetary_nebula

The Helix Nebula, NGC 7293, a planetary nebula showing a pre-WD in the center
emitting UV that excites the surrounding gas and produces inner rim cometary knots & filaments.
Distance 650 ly (200 pc), width of uncropped image 27 arcmin = 5.1 ly = 1.6 pc
(May 2003, cropped NASA HST image)
232

Stellar Clusters

Stellar or star clusters are groups of stars that are not galaxies. They are often the building blocks of
galaxies. They are very useful in testing theories of stellar evolution. It is generally safe to assume that all
the stars in a cluster formed at about the same time from similar starting compositions and with the same
distance modulus. The distance modulus can be determined by matching the main sequence of the cluster to
a main sequence calibrated in absolute magnitude, a process called by the misnomer spectroscopic parallax,
also referred to as main-sequence fitting (discussed further below).

Metallicity Z, [Fe/H], and [O/H]
In astronomy, the term metals is used (somewhat perversely!) for atoms and nuclei other than H and He.
The term metallicity is given by Z, where Z=mass fraction of nuclei having atomic number > 2. The Sun has
Z0.016 to 0.018. Recall that X is the mass fraction of H (H kg/total kg), and Y is the mass fraction for He
(He kg/total kg), and that the mass fractions X + Y + Z = 1.

232
Another measure of metallicity, expressed relative to the Sun, is [Fe/H], defined by:
e/ o

Here N are number densities (atoms per unit volume). Similar definitions apply to [O/H].
The metallicity [Fe/H] of the Sun is 0.0 as hereby defined. Here,
233

o

Stellar Populations
Stars are grouped on the basis of metallicity (given by Z or [Fe/H]. The hypothetical earliest stars, produced
soon after the big bang, consisted when first formed almost entirely of H and He. They therefore began with
no metals content, so Z = 0. These are the hypothesized metal-poor Population III stars. More recently
formed but still very old and metal-poor stars with Z0 are termed Population II. Newer stars, which had
more metals available to incorporate, have mass fraction values as high as Z0.03 and are termed metal-rich
Population I stars The Sun is in this population I.

Definitions of Open and Globular Stellar Clusters
There are two types of star clusters: globular clusters, which are tight groups of hundreds of thousands of
very old stars that are gravitationally bound, and open clusters, which are more loosely clustered groups of
stars, generally contain fewer than a few hundred members, and are often very young.
234

Open clusters, also confusingly called galactic clusters (these are not clusters of galaxies), tend to be smaller
and contain younger Pop. I stars, numbering a few thousand. They have been found in spiral and irregular
galaxies. Many are inherently unstable... There are over 1000 in the Milky Way. Open clusters generally
survive for a few hundred million years. In contrast, the more massive globular clusters of stars exert a
stronger gravitational attraction on their members, and can survive for many billions of years.
235
Open
clusters become disrupted over time by the gravitational influence of giant molecular clouds as they move
through the galaxy, but cluster members will continue to move in broadly the same direction through space
even though they are no longer gravitationally bound; they are then known as a stellar association,
sometimes also referred to as a moving group.
236

Globular Clusters GC
A globular cluster is a spherical collection of stars that orbits a galactic core as a satellite.
237
It is not a
galaxy but is tightly bound by gravity, giving spherical shapes and high stellar densities. Examples include
Messier 80. They are found within galaxies in the galactic halo. They consist of tens to hundreds of
thousands of stars. The M13 cluster like most globular clusters consists of very old Pop. II stars. .
Large collections of cluster stars are often studied in color-magnitude diagrams, plotting for instance V
[visual band magnitude] on the ordinate and B - V on the abscissa. The patterns of stars trace out various of
the stages of stellar evolution:

233
http://burro.astr.cwru.edu/Academics/Astr222/Galaxy/Structure/metals.html
234
http://en.wikipedia.org/wiki/Star_cluster
235
http://en.wikipedia.org/wiki/Open_cluster
236
http://en.wikipedia.org/wiki/Star_cluster
237
http://en.wikipedia.org/wiki/Globular_cluster

Color Magnitude Diagram for Globular Star Cluster M55, NASA, February 2001.
The Main Sequence, Blue Stragglers, Turnoff to the SGB, SGB, RGB,
and HB including Blue Giants are apparent. The E-AGB and Post-AGB are apparently not included. (NASA)
238

The position of the sharp upper turn-off from the main sequence toward the subgiant and red giant branches
is a useful index of the cluster's age. The more time that has elapsed, the lower will be the average masses of
stars remaining on the main sequence, and thus the turnoff will be toward the later spectral types. A series
of turnoff curves vs. estimated age is shown on IMA2 p. 478. A few blue stragglers may be present above and
to the left of the turnoff.
The relative number of stars at various positions in the MS versus in other evolutionary stages at a given time
(a pattern termed an isochrone) also gives indication of the cluster age.
Stars in the Hertzsprung Gap are moving rapidly through this often sparsely populated SGB location toward
the RGB stage. For similar reasons, namely the brevity of these stages, few stars may be apparent in the AGB
and Post-AGB phases as stars evolve toward the WD stage.

238
http://apod.nasa.gov/apod/ap010223.html ; see also IMA2 p. 476
Stellar Pulsation (Chapter 14)

This chapter deals primarily with pulsating variable stars (i.e., those that intrinsically and repetitively pulsate
over significant amplitudes). I can only scratch the surface of this rich and fascinating subject matter, a
professional focus for Prof. Szkody.
This chapter excludes:
eruptive variablesstars that undergo cataclysmic changes (these arise from a WD primary with a mass
transferring secondary, and include novae, dwarf novae, novalike eruptions, polars and low accretion rate
polars LARPs, and AM CVn stars)
239
or catastrophic changes (such as supernovae); stars that emit eruptive
flares, mass ejections (such as Wolf-Rayet stars), or X-ray bursts or gamma-ray bursts, and
extrinsic variable starsstars whose variability results largely from their rotation (by having prominent
starspots, polar brightness variation, ellipsoidal shapes, etc.), as well as eclipsing binary systems).
240

The International Variable Stars Index (AAVSO VSX) is a useful place to find virtually all major variable
stars listed by type and searchable by applying various criteria, though tricky to use.
241

Key Observations of Pulsating Stars

David Fabricius first noted in 1595 the variation in the brightness of o Ceti (omicron Ceti), which he termed
Mira, meaning wonderful. (Earlier observations of this stars variability are uncertain.) Its period of 11
months was found by Johannes Holwarda, making this a long-period variable (those having periods of 100 to
700 days), and it is now the prototype of the Mira Cet variable type.
242
Mira is actually a binary star system at
~90 pc, and Mira A is the much larger and variable component. Mira A is a red giant and said to be an
Asymptotic Giant Branch (AGB) star, in the thermally pulsing (TP-AGB) phase.
243
The oscillation period
averaging 333 days varies from about 328 days to about 336 days over a longer period of about 40 years.
244

The variation within a single period extends a wide range , from magnitude ~2 to ~10. The variation is due to
the star expanding and contracting, with
changes in brightness resulting.
Another pulsating star, whose variability
was discovered in 1784, is Cephei
the prototypical classical Cepheid. It
has a period of 5d 8hr 48min and
magnitude variation of about 0.7
magnitudes. Henrietta Swan Leavitt
(18681921) working at Harvard,
identified more than 2400 classical
Cepheids with periods from 1 to 50 days,
mostly in the SMC. She plotted and
observed the close relationship between
period and luminosity. A more recent
1968 graph of this Period-Luminosity
(PL) relation follows:
This graph at right shows the Period-
Luminosity relationship for classical

239
http://heasarc.gsfc.nasa.gov/docs/objects/cvs/cvstext.html
240
International Variable Stars Index:
http://www.spaceacademy.net.au/library/guides/varstar/varstar.htm
http://en.wikipedia.org/wiki/Variable_star
241
http://www.aavso.org/vsx/index.php?view=search.top
242
http://en.wikipedia.org/wiki/Mira_variable
243
http://en.wikipedia.org/wiki/Mira
244
http://arxiv.org/abs/0902.3604
Cepheid variables in MW clusters Cepheids, in the open cluster h and Persei, and in the SMC, LMC, M31,
and NGC 6822 (Barnards Galaxy). (Sandage and Tammann 1968
245
, graph for MB omitted)
According to IMA2 p. 486, this data is fitted to the equation
MV = 2.81 log10 Pd 1.43,
where Pd is the period in days and MV is the average absolute V magnitude. Sandage and Tammann state,
The data are consistent with the view that the P-L relations are the same from galaxy to galaxy in both slope
and zero point. Tighter results
are reported using the infrared H
band where extinction is lower.
An even tighter correlation arises
from incorporation of an infrared
color index J band KS band, so
that the best relation comparing H
band absolute magnitude, period P
in days, and the average color
index J KS is reported by IMA2
(p. 487) to be
o
+

This is called a Period-
Luminosity-Color (PLC) relation.
Persson et al (2004) give recent a
PLC calibration of Cepheids in the
LMC using near-infrared light
curves as follows:
246

The graph to the right is a PLC plot
for Cepheids in the Large
Magellanic Cloud. Here, the
vertical axis mag incorporates the
W(KJS) Wesenheit parameter and
(according to IMA2 p. 488) is of the
form H c(JKS). These
corrections provide the tightest
correlation with Log Period.
(Persson SE et al 2004)
(I have not yet taken time to study
the complex relationships described in these technical articles.)
Classical Cepheids, which are supergiant stars of high luminosity class Ib (i.e., a less luminous supergiant,
thus having luminosity~1000 L), provide one of the standard candles needed to estimate intergalactic
distances.

245
Sandage AR & Tammann GA, A Composite Period-Luminosity Relation for Cepheids at Mean and
Maximum Light, Astrophysical Journal, vol. 151, p.531, 1968.
246
Persson SE, et al, New Cepheid Period-Luminosity Relations for the Large Magellanic Cloud: 92 Near-
Infrared Light Curves. The Astronomical Journal 128 2239, 2004,
http://iopscience.iop.org/1538-3881/128/5/2239
The Pulsation Hypothesis for Brightness Variations
Harlow Shapley (18851972) showed that the variability of classical Cepheid stars is not attributable to their
being in binary systems, and suggested that the variability of brightness and T arises from radial pulsation of
a single star. Arthur Eddington (18821944) provided the theoretical underpinning of this theory. For
Cephei, the full excursion or change of brightness (which might be expressed as L or Mbol) is due primarily
to the ~1000 K excursion in surface temperature (Teff), with a lesser contribution to L or Mbol due to
excursion in radial size R (increased r of course increases radiating surface). R/Rmean is only about 5 to
10%, though R may be greater than 2R.in these giant stars. The spectral type varies during the pulsation
cycle from F5 at the peak to G2 at the nadir.
According to Martin Schwarschilds 1938 observed and derived data for Cephei (shown in graphs below),
247

the Teff and Mbol peaks are at almost simultaneous phases as expected (phases 0.98 and 1.00, respectively)
and occur somewhat past when the star is most rapidly expanding (when is maximally positive, at phase
0.90) and when the star passes through the minimum radius (at phase 0.92).

Variability in Cephei over one cycle:
L-R, top to bottom:
Stellar Radius r, Surface Temp Teff, Surface Velocity ,
T/T
, Surface Acceleration , and Ratios of P/P
and /
.
248

247
Martin Schwarzchild, On the Light Curve of Cephei Harvard College Observatory Circular, vol. 431,
pp.1-13, 12/1938 For exact phaes, see Table VI. [Martin Schwarzchild (1912 - 1997) is son of Karl S.]
248
ibid.
The Instability Strip and H-R Regions Where Pulsating/Variable Stars Are Found

HR Diagram showing Regions of Stellar Variability or Pulsatility.
249

(Note that spectral classes depicted here are aligned vertically,
not always the case for H-R diagrams)
Regions designated here include the following (not all of which exhibit regular pulsation)
LBV=Luminous Blue Variables (S Doradus);
YHG=Yellow Hypergiant variables;
Classical Cepheid (Type/Population I Cepheids, including Cephei);
LPV=Long Period Variables (Mira variables);
BL, W, & RV Cepheids (Type/Population II Cepheids, metal poor, incl. BL Herculis, W Virginis,
and RV Tauri stars)
RR=RR Lyrae stars;
Sct= Delta Scuti stars;
Dor=Gamma Doradus stars;
Cep=Beta Cephei stars;
SPB= Slowly Pulsating B stars (53 Persei stars);
DAV=White dwarf variables (aka ZZ Ceti);
DBV=White dwarf variables (aka V777 Her);
PG=PG 1159 stars (very hot white dwarf variables);
Flare=Flare stars (mostly class M red dwarfs).

249
image (modified by MCM) fr. http://stars.astro.illinois.edu/sow/hrd.html, fr. JB Kaler, The Cambridge
Encyclopedia of Stars, Cambridge U Press, Nov 2006
Obviously, the detailed classification of variable stars is quite complex. The types of variable stars and full
descriptions are listed here
250
. This summary mentions other variable star types not listed in this table.
Here are some of the main types of pulsating variables (plus a few ringers), listed in top-down order as shown
on the HR Diagram above (definitions and quoted material from AAVSO website (www.aavso.org) unless
otherwise noted):
Type of
Variable Star
Period
(some stars are
multiperiod or
nonperiodic)
Notes
(some edited with added abbreviations)
LBV=Luminous Blue
Variables (aka
S Doradus variables)
not regular
pulsations
... are very bright, blue, hypergiant variable stars named after
S Doradus, the brightest star of the LMC. They exhibit long,
slow changes in brightness, punctuated by occasional
outbursts in brightness during substantial mass loss events
(e.g. Eta Carinae, P Cygni). They are extraordinarily rare. The
General Catalogue of Variable Stars [GCVS] only lists 20
objects as SDor.
251

YHG=Yellow
Hypergiant variables
? not regular
pulsations
...form a class of very luminous yellow supergiants that show
strongly turbulent photospheres and high rates of mass loss...
With effective temperatures of only 40007000 K, they occupy
the cooler half of the Hertzsprung gap, which is the wide
yellow band that separates red supergiants from luminous
blue variables (LBVs or S Doradus variables)... It is now
believed that yellow hypergiants are the cores of massive stars
that have recently evolved out of a much cooler red supergiant
state...
252

Classical Cepheid:
Population I
variables
(including Cephei)
1 to 70 days

IMA2: 1-50 days
Light variations from 0.1 to 2 magnitudes. These massive
stars have high luminosity and are of F spectral class at
maximum, and G to K at minimum. The later the spectral
class of a Cepheid, the longer is its period. Cepheids obey the
period-luminosity relationship. [AAVSO] Radial mode.
LPV=Long Period
Variables (aka Mira
Variables)
30-1000 days
IMA2: 100-700
days
pulsating red giants or supergiants with periods ranging from
30-1000 days. They are usually of spectral type M, R, C or N.
There are two subclasses; Mira and Semiregular:

Mirathese periodic red giant variables vary with periods
ranging from 80 to 1000 days and visual light variations of
more than 2.5 magnitudes.

SemiregularThese are giants and supergiants showing
appreciable periodicity accompanied by intervals of
semiregular or irregular light variation. Their periods range
from 30 to 1000 days, generally with amplitude variations of
less than 2.5 magnitudes. [AAVSO] Radial mode.

250
Classification of variable stars:
http://www.aavso.org/types-variables
http://en.wikipedia.org/wiki/Variable_stars
251
http://en.wikipedia.org/wiki/Luminous_blue_variable_star
252
http://iopscience.iop.org/0004-637X/560/2/934/fulltext/53874.text.html
BL, W, & RV Cepheid
Variables:
Type/Population II
Cepheid variables,
metal poor, incl.
BL Herculis,
W Virginis, and
RV Tauri
1 to 150 days

BL Her:
1 to 5 days

W Virginis:
10 to 20 days

RV Tauri:
30 to 150 days
253

RV Tauri: These are yellow supergiants having a
characteristic light variation with alternating deep and shallow
minima. Their periods, defined as the interval between two
deep minima, range from 30 to 150 days. The light variation
may be as much as 3 magnitudes. Some of these stars show
long-term cyclic variations from hundreds to thousands of
days. Generally, the spectral class ranges from G to K.
BL Her or Herc: [of Population II Cepheids,] the short-period
stars are called BL Herculis stars, the long-period stars
generally have RV Tauri characteristicsalternating deep and
shallow minima.
W Virginis: That leaves the medium-period stars, which are
called W Virginis stars.
[all AAVSO] Radial mode (at least for W Virginis)
RR=RR Lyrae
variables
0.05 to 1.2 days

IMA2: 1.5-24hr
These are short-period (.05 to 1.2 days), pulsating, white
giant stars, usually of spectral class A. They are older and less
massive and less luminous than Cepheids. The amplitude of
variation of RR Lyrae stars is generally from 0.3 to 2
magnitudes. [AAVSO] Radial mode
Sct= Delta Scuti
variables
0.03 to 0.3 day

IMA2: 1-3 hr
...reside near the point where the instability strip crosses the
main sequence in the HR diagram. Stars in a variety of
evolutionary statesincluding pre-main sequence stars can
lie within the instability strip, so long as they have spectral
types between (roughly) F8 and A2, and luminosity classes
between V (dwarf) and III (subgiant). For stars with solar
metal abundances, this corresponds to masses between about
1.5 and 2.5 solar masses, and between 1.0 and 2.0 solar
masses for metal-poor stars. They are all short period stars,
with individual periods lying in the range of 0.03 to 0.3 day...
A few delta Scuti stars (like FG Virginis and XX Pyxidis) are
known to have a dozen or more of these non-radial modes
excited at once, and can be detected via the analysis of very
precise time-series photometry. [AAVSO] Radial & nonradial
modes
Dor=Gamma
Doradus variables
0.4 to 3 days ...class of pulsating stars, the prototype of which is the bright,
early, Ftype dwarf Doradus. These stars typically have
between 1 and 5 periods ranging from 0.4 to 3 days with
photometric amplitudes up to 0.1 mag in Johnson V [band].
The mechanism for these observed variations is highorder,
lowdegree, nonradial, gravitymode pulsation.
254

Nonradial modes

253
Periods for types of Pop II Cepheids from http://www.jstor.org/stable/10.1086/341698
254
http://www.jstor.org/stable/10.1086/316399
Cep=Beta Cephei
variables
0.1 to 0.3 days
IMA2: 3 to 7 hr
Beta Cephei variables are sometimes known as Canis
Majoris stars... But, in the General Catalogue of Variable
Stars, they are classified as BCEP ... They have spectral types
of B0-B2 III-V, corresponding to temperatures of 20-30,000 K,
i.e. they are near the top of the main sequence on the
Hertzsprung-Russell diagram. Their masses are typically 10-
20 times that of the sun. They have periods of 0.1 to 0.3 days
and, though their radial velocity amplitudes are quite large,
their visual light amplitudes are very small, typically only a few
hundredths of a magnitude... Satellite observations show
that, in ultraviolet light, where most of the flux from these hot
stars is found, their amplitudes are up to a magnitude... But
the small visual amplitude explains why the measurement of
radial velocity variations was the observational method of
choice. BW Vulpeculae, the largest-amplitude Cephei star,
has a vigorous pulsation amplitude of 200 km/s... [AAVSO]
Radial & nonradial modes
SPB= Slowly
Pulsating B stars
(aka 53 Persei
variables)
0.5-5 days
255
[Zeta] Cas is a member of an unusual group of variable stars
known as "Slowing Pulsating B" (SPB) stars. It shows a
pulsation frequency of 0.64 per day (or once every 1.56 days)
and displays a weak magnetic field with a strength of roughly
3.35 102 T, which varies with a period of 5.37 days. This
likely matches the rotation rate of the star, which, when
combined with the low projected rotational velocity, indicates
the star may be seen nearly pole-on.
256

There are a large number of Cephei stars among the
brightest stars in the sky.... Starting about [1980], there was
much interest in the newly-discovered slowly-pulsating B
stars, the 53 Persei stars, and pulsating Be stars... Non-radial
modes [AAVSO]
257

DAV=White dwarf
variables (aka
ZZ Ceti variables)
IMA2 p. 561:
100 to 1000 sec
luminosity varies due to non-radial gravity wave [g-mode]
pulsations
The DAV stars (commonly called ZZ Ceti stars) have
temperatures around 10 000 12 000 K,
258

D = degenerate, V = variable
Nonradial g-modes
DBV=White dwarf
variables (aka V777
Her)
minutes
259
g-mode oscillations driven by the helium partial ionization
zone [IMA2 p. 562]
The DBV stars have temperatures in the range ~22,000
27,000 K
260

D = degenerate, V = variable
Presumably nonradial modes
PG=PG 1159
variables (aka GW
Vir, very hot white
dwarf variables).
Divided into DOV and
PNNV stars
300 to 3,000 sec pre-degenerate, hydrogen-deficient atmosphere, surface
temperatures between 75,000 K and 200,000 K... These stars
vary slightly (510%) in brightness due to non-radial gravity
wave pulsations within themselves. They vibrate in a number
of modes simultaneously, with typical periods between 300
and 3,000 seconds.
261
nonradial modes
262

255
http://www.ster.kuleuven.ac.be/~peter/Bstars/
256
http://en.wikipedia.org/wiki/Zeta_Cassiopeiae
257
http://www.aavso.org/vsots_betacep
258
www.star.ac.za/course-resources/local/david-buckley/var2.pdf
259
Period of GD 358 given by http://adsabs.harvard.edu/abs/1982ApJ...262L..11W
260
ibid.
261
http://en.wikipedia.org/wiki/PG_1159_star
Flare=Flare stars
(mostly class M red
dwarfs including UV
Ceti)
not regular
pulsations,
flares last
seconds to
minutes
UV Ceti, however, is an extreme example of a flare star that
can boost its brightness by five times in less than a minute,
then fall somewhat slower back down to normal luminosity
within two or three minutes before flaring suddenly again after
several hours. In 1952, UV Ceti was observed flaring to 75
times its normal brightness in only 20 seconds.
263

...a variable star that can undergo unpredictable dramatic
increases in brightness for a few minutes. It is believed that
the flares on flare stars are analogous to solar flares in that
they are due to magnetic reconnection in the atmospheres of
the stars. The brightness increase is across the spectrum,
from X rays to radio waves... The best-known flare star is UV
Ceti, discovered in 1948. Today similar flare stars are
classified as UV Ceti type variable stars (using the abbreviation
UV) in variable star catalogs such as the General Catalogue of
Variable Stars... Most flare stars are dim red dwarfs,
although recent research indicates that less massive brown
dwarfs might also be capable of flaring. The more massive RS
Canum Venaticorum variables (RS CVn) are also known to
flare, but it is understood that these flares are induced by a
companion star in a binary system which causes the magnetic
field to become tangled.
264

The MW contains several million pulsating stars out of several hundred billion, so variability may be relatively
transient. The stars with substantial pulsation are not generally located on the Main Sequence, but are found
in the Instability Strip (shown above extending from classical Cepheids to RR Lyrae and Sct stars) and in
other regions noted. As star evolve through the instability strip, they begin to pulsate, and stop when they
leave it.
Ultimately, most if not all stars including the Sun
265
exhibit some degree of variability in T and luminosity,
etc., but the fluctuations tend to be relatively small and inconspicuous compared to the stars designated as
variable.
266

The Physics of Stellar Pulsation

Period-Density Relation
Radial stellar oscillations are due to sound wave resonances in the interior (IMA2 p. 491). The period of
pulsation (symbol Pi is used to avoid confusion with pressure P). The time for sound to cross the diameter
of a star is crudely estimated using adiabatic sound speed VS at IMA2 p. 492, yielding a rough estimate of the
pulsation period for the Fundamental radial mode.

where = constant average density (unrealistic) and CV = Specific Heat at constant volume and CP = Specific
Heat at constant pressure. Therefore, period is approximately inversely proportional to the square root of
mean density. For a typical Cepheid, the period this formula estimates is 10 days, which is in the right

262
http://adsabs.harvard.edu/abs/2004astro.ph..9243N
263
http://www.solstation.com/stars/luy726-8.htm
264
http://en.wikipedia.org/wiki/Flare_star
265
http://science.nasa.gov/science-news/science-at-nasa/2010/05feb_sdo/
266
http://cseligman.com/text/stars/variables.htm
ballpark. This relationship helps to explain why periods decrease as one moves down the instability strip
(toward lower luminosity) from tenuous supergiants to dense white dwarfs.

Radial vs. NonRadial Modes of Spherical Oscillations (Spherical Harmonics)
Radial Modes: The radial mode stellar pulsations are essentially standing sound waves, analogous to those of
an open-ended organ pipe. The pressure and velocity perturbations in a standing acoustic wave are 90 out
of phase. Thus at the closed end of the pipe, the velocity vanishes and the pressure fluctuations are maximal.
Similarly, at the center of a radially pulsating star, the radial velocity has a node and the pressure has an
antinode. At the open end of the organ pipe and at the surface of a star, the situation is reversed: the
pressure fluctuations vanish and those of the velocity are maximal...
267
The standing wave for each radial
mode has a velocity node at the stars center and a velocity antinode at the surface. In the fundamental
mode (first harmonic), the gases throughout the star are moving outward or inward with radial symmetry and
there is no other node than at the center. For this mode, the maximal fractional displacement of the gas from
its equilibrium position rises steadily as one moves from the velocity node at the center to the surface.
For the first overtone (second harmonic), there is a spherical velocity node at ~0.6R
, so that gas moves

inward inside this node while gas outside it moves outward, and conversely (all with radial symmetry). The
second overtone has 2 concentric spherical nodes, the third overtone has 3 concentric spherical nodes, etc.
For radial modes, most of the motion takes place at or near the surface, with lesser movement deep inside the
star. For example, the maximal fractional displacement from equilibrium for the fundamental mode at 0.5 R

is only 7% of that at the surface (1 R
). For higher overtones, the disproportion between surface versus deep

displacement is even greater. The vast majority of classical Cepheids and W Virginis stars oscillate in
fundamental radial mode. RR Lyrae and possibly LPVs like Mira stars oscillate in fundamental or first
overtone modes, sometimes combined (IMA2 p. 494).
Radial modes often produce the largest changes in a stars radius, leading to large changes in luminosity,
temperature, and radial velocity... The simplest radial mode is the fundamental mode, in which the entire
atmosphere of the star expands/contracts in unison. Nonradial modes are more complicated, no longer
radially symmetrical, as some parts of the stars surface move inwards while other parts move outwards...
Examples of nonradial pulsators include pulsating white dwarfs, Gamma Doradus stars, Beta Cepheids, and
Delta Scuti stars.
268
(For further discussion on nonradial modes, see later.)

Mechanism Powering Stellar Standing Waves
Arthur Eddington proposed that pulsating stars are thermodynamic heat engines. This is illustrated by
means of a P-V thermodynamic diagram, which shows a closed loop as one goes through a complete period. If
the integral about the complete path (the path is clockwise), the gases of the star do net P dV work
through the cycle (in a driving layer), whereas if the integral is < 0 (the path is counterclockwise), the gases
are acting as a damping layer. Oscillations will grow if the total work (sum of integrals for all the layers) > 0.
At an equilibrium value of pulsation amplitude, the total work done by all layers is zero. The driving layer
must absorb heat at around the time of maximal compression... The core of the star usually exhibits too little
pulsation amplitude to power the pulsations. (However, oscillations in a stars nuclear energy generation rate
contribute to the Eddington limit on main sequence stellar mass of about 90 M. Eddington suggested that a
layer inside the star becomes more opaque in compression, restricting the flow of energy to the surfacewith
expansion and reduced compression, the opacity diminishes and energy flow to the surface resumes at the
higher rate. This layer is referred to as Eddingtons Valve. Only in special circumstances does opacity
increase with compression, explaining the relative rarity (1:10
5
) of pulsating stars. This region was identified
as a partial ionization zone. Here, work done by the gases causes increased ionization, increasing Kramers
opacity . The opacity decreases with decreasing density. This is the mechanism.
An additional mechanism reinforces the effect. Specifically, the -mechanism is reinforced by the
mechanism in a partial ionization zone. The temperature of the partial ionization zone rises less during
compression than for the surrounding layers and the resulting stronger temperature gradient causes
relatively increased heat flow into the cooler partial ionization zone, causing further ionization. (IMA2 p. 497)

267
http://www.astro.princeton.edu/~gk/A403/pulse.pdf
268
http://www.chara.gsu.edu/~wiita/pulsating_stars2.pdf
These partial ionization zones modulate the flow of energy and directly cause the pulsation of stars (IMA2 p.
497). In most stars there are two partial ionization zones: the hydrogen partial ionization zone (H I H II
plus some He I He II) at about 1 to 1.5 x 10
4
K, and the helium partial ionization zone (He II He III) at
about 4x10
4
K.
The location of these partial ionization zones determines the stars pulsation. At hotter temperatures of 7500
K, past the hot blue edge of the instability strip (on the left), the zones are too close to the surface to be
effective in causing pulsation. At about 6500 K, pulsation is more effective and the first overtone may be
excited. In a cooler star of 5500 K, the deeper fundamental mode may be excited. However, with still cooler
surface temperatures, the cool red edge of the instability strip (on the right) is reached, causing damping by
convection (IMA2 p. 498).
The very hot Beta Cephei stars and early SPB Slowly Pulsating B stars are too hot for pulsation to be
explained by H or He partial ionization. Instead, pulsations are attributable to an iron bump in opacities
from iron ionization near 10
5
K (IMA2 p. 499).

Modeling Stellar Pulsation

This is a complex subject requiring large computer resources, but has been quite successful for Cepheids.
One approach utilizes nonlinear hydrodynamic models, as the star is not in hydrostatic equilibrium. When
conditions are optimal in the model, the partial ionization zones will drive the oscillation. In some cases the
model predicts chaotic behavior, as has been observed for some stars:

Light curves for 5 semiregular variable stars exhibiting varying degrees of chaotic variability.
Red points are actual data, curved lines are fitted to the data and smoothed.
269

269
Chaos in Stellar Pulsations:
Buchler JR, et al, Evidence for Low-dimensional Chaos in Semiregular Variable Stars, The Astrophysical
Journal, Volume 613, Issue 1, pp. 532-547, 09/2004. http://iopscience.iop.org/0004-
Another approach attempts to linearize the differential equations. This is computationally less intensive but
substantially more limited. A formula for calculating the period is illustrated (IMA2 p. 500).
For the more sophisticated but more costly approach needed for modeling some variables, the model must be
non-adiabatic, to allow heat to flow into and out of the layers, as well as non-linear. In the case where
CP/CV < 4/3, the star collapses. (I have not yet determined how valid this is and how it compares to the
Chandrasekhar limit or the SchnbergChandrasekhar limit, etc.)

Nonradial Stellar Pulsation

I have not mastered this complex subject, and have omitted many details. Recall that with radial symmetry,
parameters have the same values for a given radius r throughout the star, for all colatitudinal coordinate =
inclination =polar coordinate angle , and for all longitudinal coordinate = azimuthal coordinate angle . Of
course, there can be complex radial pulsation modes having spherical nodes and anti-nodes within the star,
but they are all radially symmetrical.
As previously mentioned, however, spherical vibration may not be radially symmetrical.
In Laplace spherical harmonics functions terminology (particularly as applied to stars), the vibratory modes at
least for and are expressed as
(sin and/or cos terms multiplied by complex exponentials in

of the form e
im
)
where
l (the spherical-harmonic degree) is a non-negative integer l 0
m (the azimuthal order) can take on 2l+1 integer values in the range l (ell) to +l
K terms are normalization constants.
A third integer parameter n describing the oscillation gives the radial orderit determines the harmonic
overtone or higher harmonic, where n = 0 corresponds to the fundamental. The value of this n parameter is
not apparent at the surface and must be inferred indirectly. The magnitude |n| of the order is then simply
the number of radial nodes in the vertical displacement eigenfunction.
270
Thus for the first overtone, n = 1
and there is one nodal sphere inside the star.
Physical parameters are given by the real parts of these complex (a + ib) functions.
The radial modes have l = 0, m=0, while n is variable. The n expresses the higher harmonics (fundamental
mode and overtones) and thus the behavior in the radial direction. All radial modes have the form

Stars may exhibit nonradial oscillations, for which l <> 0 and m can take on values in the range l (ell) to +l.
When l <> 0, the patterns for m<>0 represent travelling waves that move across the star parallel to the
equator. The nodal lines where there is no excursion of radius, define the nodes, as before. They may take
the form at the surface of colatitudinal or longitudinal circles. When m = zero, the nodal lines are parallel to
the equator (have constant ), the oscillations do not depend on longitude , and the pattern is termed zonal.
When l = m <> 0, the nodal lines are all great circles passing through the poles (constant ), the oscillations
do not depend on , and are referred to as sectoral. (IMA2 p. 503-5 and here
271
) For other combinations,

637X/613/1/532/fulltext/ Graph is fig. 1 from the HTML version of this article. (Incidentally, their search
for low-dimensional attractors classic for chaotic regimes was not successful.) and
http://en.wikipedia.org/wiki/Low_dimensional_chaos_in_stellar_pulsations
270
Stellar Oscillations and Helioseismology:
Christensen-Dalsgaard J and Gough DO, 2001, On the dipolar f mode of stellar oscillation, Monthly Notices
of the Royal Astronomical Society, Volume 326, Issue 3, pages 11151121, September 2001. This provides a
useful overview of stellar spherical harmonics in general.
See also http://arxiv.org/abs/astro-ph/0207403, Helioseismology, Rev.Mod.Phys. 74:1073-1129, 2003.
271
http://en.wikipedia.org/wiki/Spherical_harmonics
nodal lines divide up the sphere in segments, a pattern referred to as tesseral (variously defined, pertaining to
tilings with similarly shaped isohedra).
For nonradial pulsations, standing sound waves can propagate horizontally as well as radially.
P-Modes: Sound (acoustic) waves represent perturbations in which pressure P provides the restoring force
these oscillations are called p-modes. These modes have both radial and angular nodes (angular nodes are
defined for and/or ). For example, the p2 mode with l=4 and m=-3 has two radial nodes between the
center and the surface, and its angular pattern has 4 nodal lines, three passing through the poles and one
parallel to the equator. This p2 mode may be thought of as the nonradial analog to the radial second
overtone mode. (IMA2 p. 505) For nonradial p-modes, most of the motion occurs at and near the surface
(IMA2 fig. 14.17). P-modes tell us mostly about conditions at the surface of a star.

G-Modes: These Gravity modes represent gravity waves in which gravity is the restoring force. These are
produced by internal gravity waves involving sloshing back and forth of stellar gases and related to buoyancy
of stellar material. In order to have buoyancy variation, there must be vertical variation, thus l > 0. Although
peak excursion of nonradial g-modes may be at the surface (IMA2 fig. 14.18), g-modes exhibit substantial
movement in the stars interior and therefore can provide a view into the very heart of a star. They are
located deep in the solar interior, below the convection zone [of the Sun] (IMA2 p. 509). (This appears to
contradict the figure just cited.)

F-Modes: These modes represent surface gravity waves in which gravity is the restoring force. The p modes
have the higher frequencies which increase as the number of radial nodes increases, whereas the g modes
have the lower frequencies, which decrease towards zero as the number of radial nodes increases. In
addition, Cowling established numerically for a polytrope of index 3 that for each l there lies between the p
and the g modes an intermediate mode, with no radial node in the radial displacement eigenfunction. He
labeled it the f mode. Christensen-Dalsgaard tentatively states that one can label modes in a spectrum as
follows:

P-Modes have n > 0.
F-modes have n=0
G modes have n < 0
272

However, ultimately determining n appears to be more complex than this...

Acoustic Frequency for P-modes: the angular frequency of the p-mode is estimated by a computation
using the travel time from node to node:

(I am unclear if this is in radians s
-1
or Hz.) The acoustic frequency is large in the deep interior of the star,
and decreases with increasing r (IMA2 p. 506). If there is no rotation, the pulsation period depends only on
the number of radial nodes and the value of l. The nominal frequency of the p-wave is an average weighted
according to where oscillations are most energetic (the superficial regions). Star rotation introduces a
splitting of waves into antegrade and retrograde propagation, and this splitting makes it possible to measure
the rotation of the Suns interior (IMA2 p.506).

The BruntVisl (Buoyancy) frequency for g-modes: This is given as an angular frequency N with a
complex formula (IMA2 p. 508):
(
)

272
Christensen-Dalsgaard J and Gough DO, 2001, op. cit.
The buoyancy frequency is zero at the center of the star (where g=0) and at the edges of convection zones
(where A = 0), and is undefined in a convection zone. The buoyancy frequency of a g-mode is given as the
average of N across the star.
Helioseismology and Asteroseismology

This is a rapidly growing highly complex field which I have not explored in any depth and hope to learn more
when the opportunity presents.
Adjacent is a computer-generated image showing the pattern of a
p-mode solar acoustic oscillation both in the interior and on the
surface of the sun. Parameters are l=20, m=16 and n=14. Note
that the increase in the speed of sound as waves approach the
center of the sun causes a corresponding increase in the
acoustic wavelength (apparently referring to the radial or vertical
wavelength).
273

Analysis of nonradial pulsations are used in Helioseismology and
Asteroseismology. These oscillations were first seen in the Sun
in 1962 by Robert Leighton (1919-1997) et al. The solar surface
amplitude is usually quite small, with velocities of < 0.10 m s
-1

(measured with Doppler shift of spectral lines) and a luminosity
variation L/L = 10
-6
. As many as 10 million modes are
simultaneously found in the Sun (by means of Fourier analysis)!
In the Sun, five-minute oscillations of periods 3-8 minutes
have been found to be p-modes, and are located below the
photosphere within the convection zone. The relative power of
the p-modes for the Sun peak in the range of frequencies of about 2.5 to 3.5 mHz (milli-Hertz). Many p-
modes have been identified.
Differential Rotation has been analyzed with rotational frequency splitting... Below the convection zone, the
equatorial and polar rotation rates converge (IMA2 p. 511). The p-modes have also helped confirm the Suns
surface mass fraction of Helium, Y = 0.2437.
Use of g-mode data for exploring the Sun has so far been unsuccessful (IMA2 p. 512).
Solar pulsation cannot be explained by the Eddington valve mechanismalthough uncertain, it is suspected
that solar pulsation draws on the p-modes confined in the convection zone.
The Scuti variables
274
tend to pulsate in low-overtone radial modes, as well as low order p-modes, over
excursions only up to 0.8 mag with a period of minutes of a few to several hours.
SX Phoenicis variable stars exhibit similar mixed mode oscillation with shorter periods of 0.71.9 hours.
275

The rapidly oscillating Ap stars (roAp stars) have strong magnetic fields, pulsation luminosity amplitudes of <
0.016 mag, and apparently exhibit higher order low-degree non-radial p-mode oscillation. Periods of 5.6 to
23.6 minutes are reported. An oblique pulsator model is suggested.
276

273
http://en.wikipedia.org/wiki/Helioseismology
274
http://en.wikipedia.org/wiki/Delta_Scuti_variable
275
http://en.wikipedia.org/wiki/SX_Phoenicis_variable
276
http://en.wikipedia.org/wiki/Rapidly_oscillating_Ap_star
The Fate of Massive Stars (Chapter 15)

Post-Main Sequence Massive Stars That Evolve To Supernovae

These are found in the uppermost part of the H-R diagram (on the left except for the RSG), and include
Luminous Blue Variables LBV, Wolf-Rayet Stars WR, Blue Supergiant stars (BSG, OI or BI, including
Rigel), Red Supergiant stars (RSG, KI or MI, including Betelgeuse), and Of stars (O supergiants with
pronounced emission lines, no example or further details found). Very massive stars are extremely rarethe
ratio of 100 M to 1 M stars is about 1:10
6
, because their lifespans are so short, they are hard to observe,
and they just form much less frequently. However, they have a major impact on the ISM by depositing mass
and kinetic energy as stellar winds. They can also quench star formation, and their intense UV radiation
ionizes gas clouds.

Luminous Blue Variables (LBV)
These are also called S Doradus variables, as named after a star in the LMC. They are extremely rare, but
well-studied examples include Eta Carinae and P Cygni. Their high luminosity (> 10
6
L) and high
temperatures (15,00030,000K) place them in the upper left region of the HR diagram. Their masses are up
to about 125 M, at or exceeding the Eddington luminosity Limit at which hydrostatic equilibrium is lost due
to high radiation pressure, etc.
277
This may explain their high mass loss. They may also have atmospheric
pulsations instabilities. They may evolve to Wolf-Rayet stars or red supergiants before exploding into
supernovae. High rotation velocities of some LBVs may also contribute to the atmospheric instability (IMA2 p.
520).
LBVs are evolved post-main-sequence stars (IMA2 p. 520). They are in an H-R region of instability and may
live only short times after turning off the main sequence [MCM: could they never have been on it?]. Modeling
has not been perfected, and there are many uncertainties.

Eta Carinae: Eta Carinae (image below
278
): The massive eruptive variable star system Eta Carinae ( Carinae)
at 2300 pc, almost hidden in the center of the Homunculus Nebula, was fitfully variable and then underwent
a giant explosion in 1837, increasing from perhaps 2 to 4 mag to 1 mag in 1843 or 1844 (see light curve).
This Great Eruption period of variable luminosity lasted until 1856. By 1870, it had dropped back to mag 8.
It is thought to be a binary,
279
consisting of a Luminous Blue Variable (LBV) of about 120 M and a much
smaller Wolf-Rayet star of about 30 M. During the Great Eruption period in the 19th century, Eta Carinae
probably ejected a mass of 3 M, visible now as the Homunculus Nebula (occupying much of the image
below). The lobes of the homunculus are rapidly expanding and contain an elemental signature of the CNO
reaction. The LBV has an Teff 30,000K, so most of its radiation is in the UV. This is scattered and absorbed
in the nebula and re-emitted as infrared, giving a deceptively low temperature appearance. The event that
the 19th century astronomers observed was a stellar near-death experience. Scientists call these outbursts
supernova impostor events, because they appear similar to supernovae but stop just short of destroying
their star. The LBV is likely to explode into a full-scale SN in the near future... It is one of the closest stars to
Earth that is likely to explode in a supernova in the relatively near future (though in astronomical timescales
the near future could still be a million years away).
280

An interesting partial correlation of eruptive activity and periastron of the binary system is described: The
brief 1838 and 1843 events rose to peak brightness within weeks of periastron passages if the pre-1845
orbital period is ~5% shorter than at present due to the mass loss of the eruption. Each event lasted only

277
http://en.wikipedia.org/wiki/Luminous_blue_variable_star
278
http://www.spacetelescope.org/images/opo9623a/ 10 June 1996 rotated and cropped by MCM
279
http://www.nasa.gov/vision/universe/starsgalaxies/companion_star.html
280
http://www.spacetelescope.org/images/potw1208a/
~100 days... The main brightening at the end of 1844 has no conceivable association with periastron,
beginning suddenly more than 1.5 yr after periastron.
281

Eta Carinae HST composite image, consisting of UV and visible light images from the
High Resolution Channel of Hubbles Advanced Camera for Surveys, 1996.
Distance 7500 ly, ~20 arcseconds across after cropping.
282

This 2010 revision of the historical light curve for Eta Carinae zooms in on the Great Eruption
during 18221864. During this time interval, the previous light curve from Frew (2004) is in blue

281
Smith N and Frew D, 2010, infra.
282
http://www.spacetelescope.org/images/potw1208a/
(points and dotted lines), while the revised light curve... appears as black dots with error bars.
Notes about the apparent color are listed above the light curve.
The orange vertical dashes show predicted times of periastron passage ... whereas
the red hash marks are similar but with a shorter (95%) period before 1848.
The dashed red horizontal line shows the quiescent magnitude of Eta Car as it would appear
with zero bolometric correction.
283

P Cygni, another LBV, also exhibited periods of marked sudden brightening in 1600 and 1655. Other LBVs
include the Pistol Star (80 150 M), and LBV 1806-20 (130200 M
), both of which are highly luminous (see

Wikipedia entries for details).

Wolf-Rayet Stars (WR)
These are also evolved massive stars (>20 M initially) with irregular variability, and are closely related to
LBVs, though of lower variability. They were discovered in 1867 by French astronomers Charles-Joseph-
Etienne Wolf (18271918) and Georges Rayet (18391906). Around 500 have been cataloged. They have
surface temperatures of 25,000K to 100,000K, so are extremely hot. They have strong and very broad
emission lines and an unusual spectrum. They lose mass at high rates. Many are rapidly rotating.

Nebula M1-67 surrounding Wolf-Rayet star WR 124 in 2009 (ESA)
284
,
nebula diameter 11001200 arcsec, distance ~5 kpc.
ESA description of this image: M1-67 is the youngest wind-nebula around a Wolf-Rayet star, called WR124,
in our Galaxy. These Wolf-Rayet stars start their lives with dozens of times the mass of our Sun, but loose

283
Smith N, Frew D, A Revised Historical Light Curve of Eta Carinae and the Timing of Close Periastron
Encounters, 2010, http://arxiv.org/abs/1010.3719v2
284
http://www.eso.org/public/images/wr124/
most of it through a powerful wind, which is ultimately responsible for the formation of the nebula.... Ten
years ago,
285
Hubble Space Telescope observations revealed a wealth of small knots and substructures inside
the nebula. The same team, led by Cdric Foellmi (ESO), has now used ESO's Very Large Telescope (VLT) to
watch how these structures have evolved and what they can teach us about stellar winds, their chemistry,
and how they mix with the surrounding interstellar medium, before the star will eventually blow everything
away in a fiery supernova explosion... The image is based on FORS1 data obtained by the Paranal Science
team with the VLT through 2 wide (B and V) and 3 narrow-band filters.
286

There are 3 categories of WRWN, WC, and WObased on content of N, C, O as well as He and H and highly
ionized species. Numerical subdivisions (such as WC9) are used to express degrees of ionization (IMA2 p.
522). The predominance of N in WN and the trends in relative abundance from WN to WO WRs correlate with
mass loss, starting with the H envelope, and with convection from the CNO processed core, exposing the
elements nucleosynthesized in the core. If the star survives long enough, mass loss will eventually strip
away all but the oxygen component of the triple-alpha ash [i.e., ash of C and O]. (IMA2 p. 522)

Evolutionary Scheme for Massive Stars
In one version of the scheme first advanced by Peter Conti in 1976, massive stars end their lives as
supernovaethe sequences based on stellar mass are as follows:
287

M > 85 M : O LBV WN WC SN
40 M < M < 85 M : O WN WC SN
25 M < M < 40 M : O RSG WN WC SN
20 M < M < 25 M : O RSG WN SN
10 M < M < 20 M : O RSG BSG SN

IMA2 (p. 522) lists this sequence slightly differently, as follows:
M > 85 M : O Of LBV WN WC SN
40 M < M < 85 M : O Of WN WC SN
25 M < M < 40 M : O RSG WN WC SN
20 M < M < 25 M : O RSG WN SN
10 M < M < 20 M : O RSG BSG SN
Note that SN do not occur at progenitor masses M < 8 (range about 6 M to 9.5 M for this lower limit for
SNe, upper limit for WD).
288
These masses are probably ZAMS masses, but I am not certain.
Woosley et al 2002 state, Unlike lower-mass stars (M8M), however, no point is ever reached at which a
massive star can be fully supported by electron degeneracy. Instead, the center evolves to ever higher
temperatures, fusing ever heavier elements until a core of iron is produced. The collapse of this iron core to a
neutron star releases an enormous amount of energy, a tiny fraction of which is sufficient to explode the star
as a supernova...
289

285
http://hubblesite.org/newscenter/archive/releases/1998/38/image/a/ (also shown at IMA2 p. 521)
286
http://www.eso.org/public/images/wr124/
287
Massey P, Massive Stars in the Local Group: Implications for Stellar Evolution and Star Formation,
Astronomy and Astrophysics, Vol. 41: 1556, 2003
288
Williams KA, et al, Probing The Lower Mass Limit for Supernova Progenitors and the High-Mass End of
the Initial-Final Mass Relation from White Dwarfs in the Open Cluster M35 (NGC 2168), The Astrophysical
Journal, 693:355369, 2009 March 1, doi:10.1088/0004-637X/693/1/355.
289
Woosley SE, et al, The evolution and explosion of massive stars, Rev. Mod. Phys. 74, 10151071 (2002)
The evolutionary tracks for massive stars ending in SNe are shown as follows (Meynet and Maeder 2003).
These tracks begin on the left, near the mass label, probably at the ZAMS. The tracks for rotating 25 M and
40 M return on themselves in a final blueward leg of a loop. (The track for 20 M also has a blueward loop,
not shown here, IMA2 p. 537.) Mass loss is significant and is included in these models. Rotation is also
important because it drives internal mixing and enhancing mass loss (and the tracks show a corresponding
greater luminosity for rotation). Note that the most massive stars never become red supergiants. (IMA2 p. 523
and here
290
):

The Humphreys-Davidson Limit
Roberta Humphreys and Kris Davidson were the first to observe that above a certain mass of about 40 M, an
upper luminosity cutoff occurs, and by implication an upper limit to the masses of stars that could evolve to
become cooler red supergiants. It became apparent that the upper luminosity limit varies non-linearly with
temperature for hot stars. For these early type [massive] stars the maximum luminosity decreases with
decreasing effective temperature, this downward trend ceasing at around Teff = 10
4
K, after which the
maximum luminosity remains more or less constant. Humphreys and Davidson (1979) noted this boundary
and suggested that an instability leading to rapid and unsteady mass loss set in there. The observed
boundary could, in principle, be explained by steady mass loss, but only if the loss were much higher than
that which is actually observed for objects near the boundary (typically 10
-5
M per year), so it is the
unsteady mass loss in sporadic eruptions which is considered to be of foremost importance.
291

290
Meynet G and Maeder A, Stellar evolution with rotation - X. Wolf-Rayet star populations at solar
metallicity Astronomy and Astrophysics, Vol. 404, No. 3 (June IV 2003) p. 975, including graph shown
DOI: 10.1051/0004-6361:20030512
291
http://www.peripatus.gen.nz/astronomy/HerRusDia.html
Supernovae (Supernovas)

Important Historical Supernovae
SN 1006: The brightest SN in history was the SN of 1006, which reached an apparent visual magnitude of
about V = mV= 9, and left the following supernova remnant (SNR) (see image below). It was witnessed by
naked eye observers in China, Egypt, Iraq, Japan, Switzerland, and possibly North America.
292
The Chandra
XRO page states, This is a composite image of the SN 1006 supernova remnant, which is located about 7000
light years from Earth. Shown here are X-ray data from NASA's Chandra X-ray Observatory (blue), optical
data from the University of Michigan's 0.9 meter Curtis Schmidt telescope at the NSF's Cerro Tololo Inter-
American Observatory (CTIO; yellow) and the Digitized Sky Survey (orange and light blue), plus radio data
from the NRAO's Very Large Array and Green Bank Telescope (VLA/GBT; red)... This combined study of the
Chandra, CTIO and VLA/GBT observations shows new evidence for the acceleration of charged particles to
high energies in supernova shockwaves. An accompanying Hubble Space Telescope image of SN 1006
293

shows a close-up of the region on the upper right of the supernova remnant. The twisting ribbon of light seen
by Hubble reveals where the expanding blast wave is sweeping into very tenuous surrounding gas.

SN 1006 remnant at distance 7000 ly = ~2.2 kpc, 55 arcmin diameter ~20 pc.
(combined from X-ray, visible, and radio, 2008)
294

292
http://en.wikipedia.org/wiki/SN_1006
293
http://hubblesite.org/hubble_discoveries/science_year_in_review/pdf/2008/supernova_remnant_sn_1006.pdf
294
http://chandra.harvard.edu/photo/2008/sn1006c/ released 2008, images detected 2003
SN 1054: The SN of 1054 (a Type II) gave rise to a well-known SNR, the Crab Nebula (M1).
295
HST states,
The newly composed image [released in 2005] was assembled from 24 individual Wide Field and Planetary
Camera 2 exposures taken in October 1999, January 2000, and December 2000. The colors in the image
indicate the different elements that were expelled during the explosion. Blue in the filaments in the outer part
of the nebula represents neutral oxygen, green is singly-ionized sulfur, and red indicates doubly-ionized
oxygen. Pulsar PSR B0531+21 is the central star in the Crab Nebula.

Crab Nebula, at distance 6500 ly = 2 kpc, width 6 arcmin = 12 ly = 3.7 pc
(HST, composite image released 2005)
296

SN 1572, Tychos SN: Tycho Brahe witnessed naked-eye SN in 1572probably a SN Type Ia. It arose in the
MW.

SN 1604, Keplers SN: Tychos pupil Johannes Kepler also saw a naked-eye SN, namely SN 1604 in 1604
probably also a SN Type Ia and arising in the MW. It was, as of May 2012, the last SN observed that arose in
the MW and the last naked-eye SN until SN 1987A.

295
http://hubblesite.org/newscenter/archive/releases/2005/37/
296
SN 1987A: The SN 1987A, a Type II-P, is located at 51.5 kpc in the nearby LMC (50 kpc), thus not in the MW,
and arose from a blue supergiant star (BSG). It is the first naked-eye SN since Keplers SN 1604. A 1995
HST image of the region of the SN (detected in 1994), follows:

Rings associated with SN 1987A (image detected 1994, released 1995 )
297

Distance 51.4 kpc
This Type II-P SN is providing a trove of very detailed information about the evolution of SNRs. The remnants
of this SN are unusual and merit greater study than I have so far given them.
The substance of the bright inner circumstellar ring, which lies in the equatorial plane [of the binary
system] that contains the center of the SN (IMA2 p. 538), is believed to have been ejected by the progenitor
BSG star (Sanduleak -69 202) approximately 20,000 years ago. This equatorial plane is inclined 43 with
respect to the line of sight.
298
This inner ring began exhibiting variable clumps of brightening in 1996 (IMA2 p.
539, see image below), as the faster expanding SN 1987A shock front overtook and collided with the older
ring. The visible light glow from this ring arises from O III emissions (IMA2 p. 538).

297
http://hubblesite.org/newscenter/archive/releases/1995/49/image/a/ image detected in 1994, released
1995.
298
Larsson J, et al, X-ray illumination of the ejecta of supernova 1987A, Nature 474, 484486, 23 June
2011, doi:10.1038/nature10090.
Progressive development of knots or clumps of brightening in the old inner ring surrounding SN 1987A,
images excerpted by MCM from HST images.
299
Ring is ~1 ly across.
Constant star ~4 oclock is unrelated.
The two larger outer rings lie in different planes, each parallel to the equatorial plane but in front of and
behind the equatorial plane, resp., and appear offset from the center of the SN site. IMA2 p. 538 offers an
interesting hypothesis that these may be painted on formerly ejected hourglass-shaped mass by narrowly
collimated jets of radiation emanating from the wobbling companion star, but this is not universally accepted.
Alternatively, A possible explanation for the formation of the [outer] rings is that they were formed by
interaction of the [stellar] wind from the progenitor star [a BSG] with gas that was released from the star at
earlier stages.
300
Many authors showed that the rings of SNR1987A coincide with the rim of a structure
having the shape of an hourglass. The hourglass may be a Strmgren shell distorted by the inhomogeneity of
the gas or of the radiation of the star.
301
Here
302
is an animated visualization of the distribution of the 3 rings
and the associated partly-hourglass-like surface in question. Obviously, this topic will undergo continued
development.
The outer rings apparently do not consist of matter ejected in 1987, but rather may have been ejected
~20,000 years ago. Therefore, they may not be true SNRs. Further comments about this SN will be made
below.
This supernova was subluminous compared to other Type II SNe. Its peak brightness was Mbol = 15.5 (IMA2
p. 536), whereas typical Type IIs attain Mbol = 18. Analysis shows that this is because its progenitor (with
mass ~20 M) had evolved from a RSG to a BSG along a blueward loop, attaining greater density and a deeper
potential well than the usual RSG.
This SN allowed detection of fully resolved gamma rays from the decay of o
, namely the emissions at 1238

keV
303
and 847 keV, as well as 2599 keV.
304

Neutrinos from this SN were detected at several sites, arriving 3 hours before the photon burst arrived. The
plausible neutrinos detected numbered only 24. According to models, the neutrinos originated and escaped
before the shock wave reached the surface of the star to trigger the release of photons. The neutrino speed
was deduced to be very close to the speed of light, (1 10
-8
)c. (IMA2 p. 540).
The expected compact stellar remnant at the core of the SN has never been detected, nor has a companion
star surviving after the SN.
Light echoes were observed in 1988 and later for this and other SNe, some enhanced by difference imaging.
305

299
http://hubblesite.org/newscenter/archive/releases/2004/09/image/b/
300
Tziamtzis A, et al, The outer rings of SN1987A, http://arxiv.org/abs/1008.3387v3, 2010, to be published
in Astronomy & Astrophysics.
301
http://en.wikipedia.org/wiki/Supernova_1987a
302
http://i.ytimg.com/vi/wez3ElEtPT4/default.jpg ,click on Supernova SN1987a rendered with glnemo2;
see also Morris T, Podsiadlowski Ph, "The triple-ring nebula around SN 1987A: fingerprint of a binary merger."
2007, Science, 315, 1103, including Supporting Online Material DOI: 10.1126/science.1136351, which has
2 additional movies.
303
Mahoney WA et al, Gamma-ray observations of Co-56 in SN 1987A, Astrophysical Journal, Part 2 -
Letters, vol. 334, Nov. 15, 1988, p. L81-L85. Detected 1238 keV line but not 847 keV line.
304
Tueller J, et al, Observations of gamma-ray line profiles from SN 1987A, Astrophysical Journal, Part 2 -
Letters , vol. 351, March 10, 1990, p. L41-L44. Included lines at 1238 and 2599 keV.
The Classification of Supernovae
Supernovae are rare, occurring about once in 100 years in a galaxy. Based on spectra and light curves, they
are classified as follows:
Type I: lacks H lines in the spectrum (star was stripped of its H envelope)
Type Ia: have strong Si II line at 615 nm near peak light. Found in all types of galaxies.
Type Ib: no strong Si II line at 615 nm, strong He I lines at 586.6 nm, etc. Only in spirals.
May be from WN WR (Wolf-Rayet) stars.
Type Ic: no strong Si II line at 615 nm, weak or no He lines. Only in spirals. May be from WC WR
stars.
Type II: have strong H lines, such as H and H, more common than Ib or Ic, usually red supergiant
Type II-P: Reaches a (sloping) plateau during 3080 d after peak.
Type II-L: Linear decrease in light curve, no plateau.
Type II-N: have some narrow lines (this type is not mentioned in IMA2, see here.)
306

Type I: Type Ias release about 1-2x10
44
J and have absolute magnitudes MB MV 19.3 0.3. Types Ib
and Ic average 1.5 to 2 mag less bright compared to Type Ia. Type Ibs release 0.1 to about 1x10
44
J and have
absolute magnitude from 15 to 17. Type Ics release 0.1 to about 5x10
44
J and have absolute magnitude
from 16 to 22, thus often brighter than Ias or Ibs. Types Ib and Ic are found in spirals near active star
formation, suggesting they arise in massive short-lived stars. Type Is decline at ~0.065 mag d
-1
at 20 d, the
rate of decline slowing at 50 days.
307

Type II: Type II SNe have a rapid rise in luminosity. They may exhibit P Cygni profile (previously discussed).
They release 0.1 to 5x10
44
J (IMA2 p. 528 says typically 10
46
J), thus often more than Type Is. The energy
released is 1% in the form of particle KE, 0.01% as photons, and about 99% as neutrino energy. The energy
released is not due to fusion reactions, but to release of gravitational potential energy. Type II-Ps release 0.1
to 1x10
44
J and have absolute magnitude 14 to 16. Type II-Ls release 1x10
44
J and have absolute
magnitude 17. Type II-Ns release 1 to 5x10
44
J and have absolute magnitude as high as -22. (These energy
release estimates are lower than in IMA2.)
308

Pair instability SNe: This is proposed mechanism in luminous supernovas (those with absolute
magnitudes > 21), arising in very large stars with very large cores, for which the type for many is said to be
Ic, and having absolute magnitudes of up to greater than -22 (e.g., PTF09cnd): ...Extremely massive stars
with Minitial140 M (if such exist)develop oxygen cores with masses, Mcore, that exceed 50M, where high
temperatures are reached at relatively low densities. Conversion of energetic, pressure-supporting photons
into electronpositron pairs occurs before oxygen ignition and leads to a violent contraction [due to sudden
drop in radiation pressure] which triggers a nuclear explosion that unbinds the star in a pair-instability
supernova. Transitional objects with 100 M < Minitial < 140M may end up as iron-core-collapse supernovae
following violent mass ejections, perhaps as a result of brief episodes of pair instability, and may already have
been identified. Here we report observations of supernova SN2007bi, a luminous, slowly evolving object
located within a dwarf galaxy. We estimate the exploding core mass to be Mcore100M, in which case theory
unambiguously predicts a pair-instability supernova. We show that >3 M of radioactive
56
Ni was
synthesized during the explosion and that our observations are well fitted by models of pair-instability
supernovae...
309

305
Light Echoes from Supernova 1987A:
http://www.eso.org/public/news/eso8802/
http://apod.nasa.gov/apod/ap060125.html
http://universe.gsfc.nasa.gov/conferences/supernova1987a/Crotts.pdf
306
http://en.wikipedia.org/wiki/Supernova
307
ibid.
308
ibid.
309
Pair-Instability SNe:
Gal-Yam A., et al, Supernova 2007bi as a pair-instability explosion, Nature Vol 462, 3 December 2009,
doi:10.1038/nature08579, http://www.nature.com/nature/journal/v462/n7273/abs/nature08579.html
Quimby RN, et al, Hydrogen-poor superluminous stellar explosions, Nature 23 June 2011 Vol 474 , p. 487,
doi:10.1038/nature10095
Gai-Yam A, Super Supernovae, Scientific American 306, 44 - 49 (2012)
doi:10.1038/scientificamerican0612-44
Core Collapse SNe
These occur in Types Ib, Ic, and II SNe. (Type Ias are fundamentally different, and discussed in Chapter 18.)
The mechanism of core-collapse SN are similar in these 3 types. In stars > ~8 M, the final stage of stellar
evolution is not a planetary nebula but a SN.
The details of the catastrophic collapse of the iron core leading to a SN (see IMA2 p. 530 to 534) are complex
and not fully understood. Computer simulation is difficult and strongly sensitive to initial conditions
Briefly, the He-burning core builds up C and O, eventually igniting C burning. An onion-like structure of
concentric shells develops in the deep interior or corefor each element involved, an non-burning layer
enriched in an element overlies a burning enriched layer of the same element. From the core outward, the
layer pairs are enriched in Fe (the Fe core), Si, O, C, He, and, in the envelope, H/He. Elements heavier than
e
are created endothermically and cannot contribute to the stars luminosity. Each of the elements
named, other than Fe, i.e., from He to Si, undergoes fusion and burns in a shell of thermonuclear reactions
(oxygen burning, carbon burning, etc.) Progressing inward, each of the successive burning layers,
generates less energy than the previous burning layer outside it, and is of shorter duration than the previous.
Core H burns10
7
y, core He burns10
6
y, core C burns 300 y, core ) burns 200 d, and core Si burning lasts
only 2 days. The high core temperature causes energetic photons to begin to destroy heavy nuclei (by
photodisintegration), leading to conversion of some of the Fe back to He and H ions. This is highly
endothermic, and removes thermal energy from the gas, reducing its pressure, and this pressure is needed to
support the core against collapse. The core has a mass of 1.3 M to 2.5 M, depending on the stars starting
(ZAMS) mass.. At this point, the T is about 8x10
9
K. The free electrons that contribute to needed degeneracy
pressure are captured by the protons and heavier nuclei (by electron capture). The electron capture reaction
is a beta decay, and releases neutrons and much energy in the form of neutrinos that escape. The
photodisintegration and electron captures have the combined effect of greatly and suddenly reducing the
electron degeneracy pressure. The inner catastrophic iron core suddenly collapses at speeds of up to
70,000 km s
-1
(0.2c), and the core is strongly compressed (details get a little confusing). The outer layers are
no longer supported by the core. The inner core suddenly stiffens up due to Pauli exclusion principle
applying to compression of neutrons (which are also Fermions obeying Fermi-Dirac statistics), and this
causes a bounce or rebound, sending pressure and shock waves outward toward the infalling outer core.
Even neutrinos have difficulty escaping in the high density that results, causing heating in this layer.
Eventually, if the star mass and other parameters are right, the shock wave reaches the surface and blows off
the envelope. (Shock waves that are dissipated before they reach the surface abort the formation of a SN.)
The envelope is driven outward by the shock. The KE of the ejected material is about 10
44
J, the photons
carry away 10
42
J (causing a peak luminosity of 10
9
L), and the release of neutrinos carries off about 10
46

W (!)
In summary, the key events are the catastrophic collapse of the iron core, the generation of a shock wave, and
the ejection of the stars envelope.
The differences among Types Ib, Ic, and II arise from differences in envelope makeup and the amount of
radioactive material synthesized. They may also relate to which type of WR star precedes them.

Stellar Remnants
After a core-collapse SN, the inner core may become a neutron star (for stars with MZAMS < 25 M) or a black
hole (for larger stars). (IMA2 p. 534) However, the star may simply be blown apart without leaving a stellar
remnant. Type Ia, though leaving a supernova remnant, do not leave a neutron star stellar remnant,
310

whereas core-collapse SN dobut I am unsure whether there are any Type Ia exceptions to this.)

Ionization and Recombination
The light curves exhibited after the peak of a SN explosion reflect the recombination of ions and electrons
that were split apart (ionized) by the shock. The increased ionization causes greater opacity of the outer
envelope, preventing initial escape of photons. The gradual recombination of ions and electrons is

310
http://chandra.harvard.edu/xray_sources/supernovas.html
accompanied by increasing transparency and gradual emission of energy from this recombination at an
effective temperature of about 5000K. (IMA2 p. 534)

Generation of Radioisotopes
In addition to energy release from
recombination, the presence of recently
synthesized short-lived radioisotopes,
synthesized during the passage of the shock
front through the star, also contributes to the
light curve. The plateau seen with a Type IIP
SN is in part attributable to energy deposited in
the ejecta from radioactive decay of
, which
has a half-life t1/2 of only 6.1 days. (IMA2 p.
534)
This graph
311
depicts the light curve observed
for SN 1987A extending to day 1444. The time
to maximum light was slower than usual for
Type IIs, taking 80 days to attain a somewhat
weak peak of Mbol = 15.5 (IMA2 p. 536). The
solid upper line from day 1 to 500 represents
bolometric flux based on the U to M bandpass
integration. The dotted lines represent
estimates of energy deposition from the four
radionuclides shown according to the expected model proportions (except that o
is increased to 5 times
that expected based on the solar ratio of Fe-57/Fe-56, in order to better match the observations). Note that
the article quotes decay times in terms of e-folding decay times, not half-lives. After the rapid initial falloff
just past the peak (which is contributed to by rapid decay of
(t1/2 = 6.1 days), the rate of decline slows

(reflecting relative depletion of the
. The curve now follows the decay curve of o
(t1/2 = 77.2 days).

Eventually, at about 1000 days, this isotope also becomes depleted, and a further slowing of the decay
occurs, now probably following the radioisotope decay curve for o
(t1/2 = 272 days). See the article for

estimates of the amounts of these radionuclides generated. Characteristic X-ray and gamma ray emissions
from the decay chain of o
were measured for the first time after SN 1987A, namely lines at 847 keV and
1238 keV (cited earlier).
Supernova Remnants (SNRs)
These are common, and include the Crab Nebula from SN 1054, the SN 1006 SNR, the SN 1572 SNR, and the
SN 1604 SNR. The ejected material is travelling at up to 0.1c (30,000 km s
-1
), and a strong shock wave is
formed defining the leading edge of the SNR as it excites and ionized the ISM. Some of the radiation may be
in the form of synchrotron radiation indicating relativistic electrons spiraling around magnetic field lines, fed
by ongoing energy emission from a pulsar stellar remnant. The term SNR applies to ejected material, not to a
stellar SN remnant such as a NS.
Chemical Abundances and SN Nucleosynthesis
A chart of solar system number abundances of the elements has been previously provided. The textbook
(IMA2 p. 541) shows a similarly plotted graph for the relative abundance of the Suns photosphere. H and He
are much more abundant there, and are found in a 10:1 ratio (by numbers). Li, Be, and B are
underabundant and can be easily destroyed in collisions with protons. Destruction of Be requires a higher
temperature than Li. The Suns convection zone transports surface Li, Be, and B into the interior. The Sun
has an unexpectedly low Li, a finding that is inconsistent with current solar models. The elements C, O, Ne,

311
Suntzeff NB, et al, The energy sources powering the late-time bolometric evolution of SN 1987A,
Astrophysical Journal, Part 2 - Letters , vol. 384, Jan. 10, 1992, p. L33-L36,
http://dx.doi.org/10.1086/186256
Mg, and Si are relatively more abundant, due to their nuclei being multiples of alpha-particles (modified from
IMA2 p. 542 by excluding N). Core-collapse SN generate much O (whereas Type Ia SN generate much Fe).

S-Process and R-Process Nucleosynthesis
Elements with higher Z are harder to penetrate by protons due to a high Coulomb potential barrier, but
neutrons suffer no such repulsive barrier. Once a neutron combined with a nucleus, raising the atomic mass
number A to A + 1, the new nucleus may undergo beta decay. This process emits an electron, a neutrino, and
a gamma ray, leading to an increase of atomic number from Z to Z+1.
When beta decays are relatively rapid relative to N captures and the neutron captures are rate limiting, the
process is called an S-process (slow-neutron-capture process), and these tend to be present in normal
stellar nucleosynthesis and to yield stable nuclei. The S-process produces approximately half of the isotopes
of the elements heavier than iron (the other half from R-Processes).
When beta decays are relatively slow and rate limiting compared to more rapid neutron captures (such as
immediately after a core-collapse SN, when there is an extremely high neutron flux), the process is called an
R-process (rapid-neutron-capture process). These occur commonly in SN reactions, and result in relatively
N-rich nuclei.
The two types of processes therefore play important roles in chemical evolution. They account for the
abundances of elements with A > 60.

Gamma-Ray Bursts GRBs and Associated X-ray Emissions

GRBs were first noted with the Vela satellites launched in the 1960s to monitor the compliance of Soviet and
other nuclear nations with the 1963 Partial Test Ban Treaty through detection of gamma rays. In 1967, the
Vela 4 and Vela 3 satellites first detected a flash of gamma radiation unlike a nuclear weapons signature... As
additional Vela satellites were launched with better instruments, the Los Alamos team continued to detect
inexplicable gamma-ray bursts. By analyzing the different arrival times of the bursts as detected by different
satellites, the team was able to rule out a terrestrial or solar origin. The discovery was declassified and
published in 1973.
312
GRBs are not the same as Soft Gamma Repeaters SGRs (described later).
Currently active NASA-related satellites include: the High Energy Transient Explorer HETE-2 (2000); ESAs
INTErnational Gamma-Ray Astrophysics Laboratory INTEGRAL (2002); Swift (2004); and Fermi Gamma Ray
Space Telescope (formerly Gamma-Ray Large Area Space Telescope GLAST, 2008). The Compton Gamma Ray
Observatory CGRO (19912000) was a NASA mission, one of the Great Observatories in space, and included 4
instruments including Burst And Transient Source Experiment BATSE. (Graphs of bursts from this platform
are shown at IMA2 p. 543.) NASAs Gamma-ray Coordinates Network
313
keeps track of these events and
provides rapid notification to subscribers.
These bursts are detected with current instruments about once per day, in energies ranging from 1 keV (soft
X-rays) to many GeV. NASA describes two types: (1) short-hard GRBs (< 2 s, averaging 0.3 s) associated
with neutron starneutron star mergers or neutron starblack hole mergers but not with SNe; and (2) long-
soft GRBs (> 2 s, averaging 30 s, and having less energy in gamma rays than short-hard)sometimes
associated with SNe or even hypernovae or collapsars.
314
Overall, GRBs last over a range of 0.01 s to 1000 s.
The integrated energy flux per unit area of detector is expressed as fluence, and has measured as great as
10
-7
Jy m
-2
.

312
Klebesadel RW et al, Observations of Gamma-Ray Bursts of Cosmic Origin, Astrophysical Journal, vol.
182, p.L85, DOI: 10.1086/181225.
See also http://en.wikipedia.org/wiki/Vela_%28satellite%29
313
http://gcn.gsfc.nasa.gov/about.html
314
http://imagine.gsfc.nasa.gov/docs/science/know_l1/grbs_explosion.html
GRBs have been found to have an isotropic angular distribution across the celestial sphere, and therefore
must arise extragalactically (see graphs of all-sky distribution of GRBs and output of GRB 970228 below).
Because of their great distances yet strong signal, GRBs are among the most energetic events in the universe,
comparable in energy to core-collapse SNe (IMA2 p. 547).

All-Sky Map Showing Isotropic Distribution of 2704 GRBs detected by BATSE,
a part of CGRO, over a 9 year period (NASA, 2007)
315

Some long-soft GRBs are caused by SNefor example, GRB 980425 in 1998 was shown to be associated with
a distant unusually strong SN, SN 1998bw,
316
thought to be a Type Ib or Ic (IMA2 p. 547 and here
317
).
However, although the GRB was atypically weak, the SN was unusually strong for a Type Ib/c, estimated at
2 to 6 x 10
45
J. Another supernova associated with a GRB (GRB 030329) was SN 2003dh, although these
correlations appear to be rare. Most long-duration GBRs are not associated with observable SNe. Yet the
indirect evidence ... supports the consensus view that most long-duration GRBs arise from the death of
massive stars.
318

315
http://heasarc.gsfc.nasa.gov/docs/cgro/batse/
316
http://www.ing.iac.es/PR/newsletter/news2/gamma.html
317
http://www.nature.com/nature/journal/v395/n6703/full/395670a0.html
318
Woosley SE and Bloom JS, The SupernovaGamma-Ray Burst Connection, Annual Review of Astronomy
and Astrophysics, Vol. 44: 507-556, 2006, DOI: 10.1146/annurev.astro.43.072103.150558

The graph
319
to the right illustrates the
gamma ray output for a strong GRB (CGRO
Flare #3595) occurring on 31 January
1993. It had a peak count rate of 1956
counts s
-1
over 2000 cm
2
in the range 25 to
100 keV and lasted about 6 min:

Rapidly decaying X-ray emissions related to the 1997 GRB 970228
320

(not the same GRB as graphed just above).

319
ftp://umbra.nascom.nasa.gov/pub/batse/batsegifs/batse_3595.gif
320
http://www.mpe.mpg.de/~jcg/grb970228.html
The GRB of 2/28/1997, which lasted < 135 s as GRB 970228, was detected by the BeppoSAX satellite (in
orbit 19962003), initially as a gamma ray burst. Subsequent X-ray detection (as 1SAX J0501.7+1146) was
done 8 hours later. The MECS instrument detects 2-10 keV (soft X-rays). Correlation with optical
counterparts localized the burst to a distant galaxy (though no SN was observed). The first image above
321
is
of X-rays at 8 hour after the GRB and shows the pattern of marked increased counts in a region localized to
the constellation of Orion. The second image is 2 days later, and shows a marked fall-off in X-ray count rate.

Mechanism of Long-Soft GRBs
The Long-Soft GRBs are thought to result from relativistically beamed jets (previously discussed). Highly
directional jets reduce the amount of total energy that must be emitted to produce the GRB observed. For a
Lorentz factor of = 100 (velocity= 0.99995c) and assuming we are in the relativistically beamed jet, the
required energy output would be reduced by 1/
2
or 1/10,000 (IMA2 p. 548). According to the hypernova or
collapsar model of Woosley and Bloom (2006), a sufficiently massive star (such as a WR) undergoing a SN may
leave a black hole with a collimating debris-laden disk and a magnetic field which together produces a highly
relativistic beamed jet. This emerging jet interacts with overlying infalling envelope material, producing
bursts of gamma rays. (Some details omitted.) The following image depicts a model of a jet emerging from a
black hole in the center of a 15 M WR star undergoing a supernova (Woosley and Bloom, 2006):

321
Gamma-ray burst 970228:
http://www.asdc.asi.it/bepposax/first/grb970228.html including X-ray image shown
http://www.sron.nl/divisions/hea/sax/htdocs/science_grb970228.html describes the GRB
Cosmic Rays and Solar Energetic Particles

Cosmic rays were discovered by Austrian physicist Victor F. Hess (18831964) in 1912, using a precision
electroscope
322
carried aloft in a high-altitude balloon. He showed that the level of radiation decreased up to
an altitude of about 1 km, but above that the level increased considerably, with the radiation detected at 5 km
about twice that at sea level. His conclusion was that there was radiation penetrating the atmosphere from
outer space, and his discovery was confirmed by Robert Andrews Millikan in 1925, who gave the radiation the
name cosmic rays.
323

The term Cosmic ray is a misnomer because primary cosmic rays initially consist of single charged particles
(mostly protons), not rays of radiation (photons). Most are protons, a much smaller amounts are alphas, and
most of the rest are nuclei of higher Z including Fe, Ni, and others up to uranium U. (See graph below for
more precise relative proportions of various nuclei over a range of energies.
324
) Small numbers of high energy
electrons, positrons, muons, pions, and lambda baryons, possibly antiprotons, etc. also qualify as primary
cosmic raysthey are more commonly produced as secondary cosmic rays. Cosmic rays are variously
defined, but the following may be suitable: The cosmic radiation incident at the top of the terrestrial
atmosphere includes all stable charged particles and nuclei with lifetimes of order 10
6
years or longer.
Technically, primary cosmic rays are those [stable charged] particles accelerated at astrophysical sources
and secondaries are those particles produced in interaction of the primaries with interstellar gas.
325
The
defining lower energy boundary (if any) for primary cosmic rays is unclear to me, but appears to be about 10
7

eV (0.01 GeV)or 10
8
eV (0.1 GeV). In addition, there seems to be some uncertainty as to whether very high
energy gamma rays can ever be considered as primary cosmic rays (probably not).

Primary cosmic rays lead to a shower or cascade of secondary cosmic ray particles, neutrinos, and photons,
as a result of collisions typically in the upper atmosphere.

A single cosmic ray particle can have energy in excess of 10
20
eV (16 J)these are called ultra-high-energy
cosmic rays (UHECR). One cosmic ray particle, the Oh-My-God Particle detected by ground-based Flys Eye
II,
326
was probably a proton. It had energy of 3.2x10
20
eV or 51 J, equivalent to the KE of a 142 g hardball
baseball traveling at 96 km/h.) This particle was moving so fast that in its rest frame, the time for it to travel
from the Virgo Cluster would be only 1.15 hours.
327
Such extremely high energy cosmic rays are not deflected
by the magnetic field. However, the energy of most cosmic rays is between 10
7
eV and 10
10
eV (see graph).

Most primary cosmic rays arise outside the Solar System, but the Sun is the source of relatively low energy
cosmic rays (Solar energetic particles SEP or solar cosmic rays) arising from solar flares, recombination
events, and CMEs.
328
They consist of protons (plus generally lower energy electrons), and have energies of a
few tens of keV to rarely above 10 GeV (10
10
eV), and many are excluded of diverted to the poles by the Earths
magnetic field.
329
Typical peak SEP energies are about 10
7
eV (IMA2 p. 551)

322
http://en.wikipedia.org/wiki/Electroscope
323
http://en.wikipedia.org/wiki/Victor_Francis_Hess
324
http://pdg.lbl.gov/2011/reviews/rpp2011-rev-cosmic-rays.pdf an excellent overall review from the
Particle Data Group by K. Nakamura et al, 2011
325
http://pdg.lbl.gov/2011/reviews/rpp2011-rev-cosmic-rays.pdf
326
http://www.cosmic-ray.org/reading/flyseye.html
327
http://www.fourmilab.ch/documents/OhMyGodParticle/
328
Solar Energetic Particles SEPs:
http://helios.gsfc.nasa.gov/sep.html
http://solarphysics.livingreviews.org/open?pubNo=lrsp-2006-2&page=articlesu15.html
329
http://www-spof.gsfc.nasa.gov/Education/wsolpart.html
The graph to the right shows the steady
decrease in the numbers of cosmic rays
(expressed in particles per (m
2
sr GeV sec))
with increasing kinetic energy, over the
range of about 10
8
eV (1.6 x 10
-10
J) to the
current record-holder of 3.2 x 10
20
eV (51
J))
330
The far left of the graph includes the
upper range of Solar Energetic Particles.
The knee region is associated with SNe.
The rest energy of a proton is almost 10
9

eV, and is presumably not included in the
particle energies shown in the graph.

Ultra-high energy cosmic rays are
hypothesized to arise from Larmor gyrating
particles repeatedly encountering the
advancing shock wave front of a SN.
Particles greater in energy than 10
15
eV
experiencing the Lorentz force arising from
the typical magnetic field of the ISM are not
likely to remain bound in the dimensions of
a SN remnantthey are therefore
candidates for becoming extragalactic
cosmic rays (IMA2 p. 552) Other causative
mechanisms may apply to cosmic rays with
energies much above the knee at 10
15
eV.
For instance, at the ankle of the graph near
10
19
eV, it is suggested that acceleration in
the vicinity of neutrons stars or black holes
may be involved (IMA2 p. 553)

330
http://www.physics.utah.edu/~whanlon/spectrum.html, data apparently taken from
Cronin J, Gaisser TK, Swordy SP. Cosmic Rays at the Energy Frontier. Scientific American 276, 1, 44, 1997.
The following is a recent 2011 graph showing numbers of primary cosmic ray nuclei versus KE for multiple
nuclear species. Note that energies for these nuclei begin above 10
8
eV and that protons (H) predominate,
followed by small amounts of He, and much lower amounts of C, O, and Ne. N is not included in this graph,
of uncertain significance.
331

331
http://pdg.lbl.gov/2011/reviews/rpp2011-rev-cosmic-rays.pdf
The Degenerate Remnants of Stars (Chapter 16)

White Dwarfs (WDs)

First Discovery
The first white dwarf discovered was in the triple star system of 40 Eridani, which contains the relatively
bright main sequence star 40 Eridani A, orbited at a distance by the closer binary system of the white dwarf
40 Eridani B and the main sequence red dwarf 40 Eridani C. The pair 40 Eridani B/C was discovered by
William Herschel on 31 January 1783.
332

The next to be discovered was Sirius B. After measuring the distance to 61 Cygni by stellar parallax,
Friedrich Wilhelm Bessel in 1838 measured the distance to Sirius. It is the brightest star in the sky, and is
actually a binary star system consisting of a white A1V MS star and a faint white dwarf at about 2.64 pc.
Bessel detected a wobble of Sirius (actually Sirius A) and deduced in 1844 the presence of an unseen
companion (Sirius B). The wobble is due to the elliptical orbit with period 50 y of Sirius A about the shared
barycenter. Alvan Graham Clark in 1862 actually observed the predicted companion, which was near
periastron. Current estimate of the Sirius B mass is 1.053 M. The pair differ greatly in luminosity:
LA = 23.5 L while LB = 0.03 L. Spectroscopy in 1915 revealed that Sirius B is an unexpectedly hot blue-
white star with teff = 27,000 K, a white dwarf. By Stefan-Boltzmann Law, the radius was estimated at only
0.008 R, its average density estimated at an enormous 3x10
9
kg m
-3
, and its surface gravitational
acceleration estimated at a very intense 4.6x10
6
m s
-2
. The spectrum from the surface atmosphere exhibits H
absorption lines with pressure/collisional broadening characteristic of DA WD stars. (H is otherwise depleted)
in these WDs, IMA2 p. 557 and p. 270.

Classification of White Dwarfs
WDs occupy the lower right corner of the H-R diagram. Their spectral type is D for dwarf or degenerate. They
are subdivided first by spectroscopic features, using a second letter:
DA: WDs having strong H Balmer absorption lines, typically pressure-broadened
DB: WDs having He I absorption lines, no H lines
DO: WDs having He II absorption lines, no H lines
DC: WDs having no absorption lines, only a featureless continuum
DQ: WDs having carbon features and spectral lines
DZ: WDs having metals features and spectral lines
DX: WDs insufficient or unclear data preventing classification
A hybrid WD might be designated DAB, meaning having H and He I absorption linesalso, DAO, DAZ, or DBZ
may be employed.
An added V means variable. For instance, DAV are pulsating white dwarfs (aka ZZ Ceti).
A temperature index number may be added, such as DA2 (as with Sirius B). The temperature index number
is computed by dividing 50,400 K by the effective temperature.
333

A luminosity class of VII is rarely applied.
The masses of DA WDs peak at a value of 0.56 M and 80% lie within 0.42 to 0.7 M (IMA2 p. 560).

332
http://en.wikipedia.org/wiki/White_dwarf
333
ibid.
The SDSS identified more than 7500 WDs that were included in the SDSS 2005 DR4 (see histograms to
follow, all quotes and images in this table come from this)
334
article):

WD DAs: For the DAs, The mean mass for the DA
stars brighter than g = 19 and hotter than Teff =
12,000 K is
MDA = 0.593 0.016 M.
In this graph, the main Gaussian fit peak is shown in
red and smaller additional Gaussian fits are shown in
other colors. 69% of DAs are enclosed by the red
Gaussian curve, which extends from about 0.45 to
about 0.7 M. (DR1 is SDSS data from earlier dataset
published in 2004, PG is from Palomar Green survey
in 2005.)
WD DBs: For the DBs with g [green band apparent
magnitude] < 19 and Teff > 16000K, the average
mass was found to be
MDB = 0.711 0.009 M.
It appears the mean mass for DB white dwarf stars
may be significantly larger than that for DAs. We
also report the highest mass white dwarf stars ever
found, up to 1.33M. The histograms are actual
data, the 4 Gaussian fits approximate portions of
their distributions. No prior studies included.

Central Conditions and Origin
The core or central pressure of Sirius B Pc is crudely estimated to be about 3.8x10
22
N m
-2
(IMA2 p. 559),
which is more than a million times that of the Suns core P. The core temperature Tc of Sirius B is crudely
estimated at around 7.6 x 10
7
K. Despite the high temperatures, it is deduced that thermonuclear reactions
are not contributing to WD luminosity, and that their centers are incapable of fusion reactions at the
estimated Tc and Pc. The core and much of remainder of most WDs (other than the atmosphere) are thought
to be primarily ionized C and O nuclei.
WDs arise from intermediate mass progenitor stars of less than 8 or 9 M , forming in the stellar cores in the
AGB or post-AGB or after a planetary nebula stage that by shedding outer layers exposes the core.

334
Kepler SO, et al, White dwarf mass distribution in the SDSS, Monthly Notices of the Royal Astronomical
Society, Volume 375, Issue 4, published online: 12 FEB 2007, DOI: 10.1111/j.1365-2966.2006.11388.x
Small WDs may retain unburned He and become helium WDs. Certain larger WDs of 8 to 10 M may have
fused C but not Ne, and consist largely of oxygen, neon, magnesium, and other intermediate mass elements.
These are called oxygenneonmagnesium WDs, and are detected by ONeMg novae).(IMA2 p. 560 and
here
335
)

Surface Conditions and Spectra
As mentioned above, the spectral characteristics are determined by the composition of the atmospheres.
Because of the extreme gravity, the H if any is present rises to the very top and heavier elements sink toward
the core stratified by their density. This means the outer layer of H, which gives the DA its H spectrum, may
be very thin and overly a thin layer of He, which overlies the O-C core. This stratification takes place quickly
over about 100 years. Convection may mix He into the H atmosphere, diluting out the H and changing the
type from DA to DB.

Pulsating WDs (DAV=ZZ Ceti variables, DBV=V777 Her variables, DOV)
DAV WDs (ZZ Ceti variables) have Teff of 10,000 to 12,000K, and periods of 100 to 1000 sec (as previously
discussed). They pulsate in non-radial g-modes that affect temperature, and that affect the mean intensity of
light emitted.
336
The almost entirely horizontal displacement of g-modes is compatible with the intense
gravitational pull, and little radial change occurs from pulsation. It is the hydrogen partial ionization zone
that drives the pulsations of DAVs. For the hotter and more luminous DBVs (V777 Her variables), the helium
partial ionization zone is responsible). The even hotter DOVs are found in the H-R diagram at the transition
from the very hot PNNV stage (planetary nebula nuclei variable) to WDs, IMA2 p. 562 and here
337
) These stars
oscillate in multiple modes simultaneously, a topic of great current research interest. The variation in light
output is small.

The Physics of Degenerate Matter

Normal thermal gas pressuregiven for an ideal gas by NkT/V, where N=number of particlesand radiation
pressuregiven by (T
/c)are entirely inadequate to support a WD against complete collapse. The

mechanism that does provide the support is electron degeneracy, which was deduced by Ralph H. Fowler
(18891944) in 1926, working with Paul Dirac (only one year after Wolfgang Pauli published his paper on the
Pauli Exclusion Principle). The Exclusion Principle requires that the total wave function for two identical
fermions is anti-symmetric with respect to exchange of the particles, or in simpler terms, that no two fermions
can occupy exactly the same quantum state. For electrons in atoms, there can be no two with the same n, l,
ml, and ms. For any 2 electrons attempting to have the same quantum numbers, the probability amplitude
resulting from adding the two parts of the anti-symmetric wave function is simply zero.
338

For fermions like electrons in a confined region such as a box, the wave functions are de Broglie standing
waves described by three quantum numbers, the components of momentum in 3 directions. Electrons can
have spin quantum number ms = 1/2, a fourth quantum number. No two electrons can have exactly the
same 4 quantum numbers. As the temperature and therefore the thermal pressure decline, electrons fill up
the lower unoccupied quantum states. A completely degenerate gas is one in which all of the lower energy
states are occupied and none of the higher energy states are occupiedthis occurs only at 0 K and can be
expressed by F. (IMA2 p. 563) For higher temperatures, some of the states with energy F will be
vacant, and some electrons will occupy states with energy > F. Here, F is the Fermi Energy, the maximum

335
http://en.wikipedia.org/wiki/White_dwarf
336
Winget DE, Asteroseismology of white dwarf stars, Journal of Physics: Condensed Matter Volume 10
Number 49 1998, doi:10.1088/0953-8984/10/49/014
337
http://adsabs.harvard.edu/abs/2007arXiv0705.2115S , Solheim et al, 2007
338
Pauli Exclusion Principle:
http://en.wikipedia.org/wiki/Pauli_exclusion_principle
http://hyperphysics.phy-astr.gsu.edu/hbase/pauli.html
electron energy found in a completely degenerate gas at 0 K. For higher temperatures, F is the energy at
which 50% of the states for that energy are occupied by electrons.
By applying the relation between momentum and de Broglie wavelength, p=h/, and the kinetic energy
relationship to momentum, = p
2
/2m, a formula for Fermi energy is derived:
/

where m is the electron (or other Fermion) mass, and n is the number of electrons per unit volume. (IMA2 p.
565)
Despite the high temperatures of a WD, it is a good approximation at such high densities to assume complete
degeneracy for the calculation, so that all but the most energetic electrons have F...
The condition for degeneracy becomes
T
<

/

(assuming Z/A, the ratio of protons to nucleons for the gas, is 0.5). The smaller the LHS is compared to (a
quantity derived from constants and the composition of the gas), the more degenerate the gas is.
Applying this formula to standard models of the Suns core, the LHS is 5500. The Sun has very low electron
degeneracy, supplying only a few tenths of a percent of the central pressure (IMA2 p. 566). However, as it
evolves along the SGB it will acquire much higher degeneracy, which peaks at the He core flash and the TP-
AGB. (IMA2 p. 567)
In contrast, applying this formula to values for Sirius Bs core, the LHS is only 37, much less than 1261 in
the same units, and complete degeneracy is a valid assumption.
As electrons are forced into more tightly defined positions, their momenta increase (by Heisenberg
Uncertainty). The rising momenta contribute a temperature-independent pressure, the degeneracy pressure,
for which the derived final formula is (IMA2 p. 569)
[(
]
/

where me is mass of an electron, mH is mass of hydrogen (almost the same as mass of a proton), Z is average
number of protons per nucleus, A is average number of nucleons per nucleus. It is this degeneracy
pressureto which is added a much smaller thermal gas pressure and radiation pressurewhich maintains
hydrostatic equilibrium against gravity in a white dwarf.
For the extreme relativistic limit, where electron speeds are close to c, IMA2 p. 570 gives an alternate formula
for degeneracy pressure:
[(
]
/

Comparing the Sun with the WD Sirius B, I calculate (using formulas from IMA2, p. 569, 291, and 295 and
assumed values shown) the following (all units are MKS):

Result
N m
-2

Z / A Mean MW
incl. electrons
Core Density
Sun Core Gas Pressure 2.74E+16 ? 0.723
339

uncertain estimate
152700
Sun Core Radiation P 1.53E+13
Sun Degeneracy P 2.17E+15 0.5

339
http://iopscience.iop.org/0004-637X/670/1/872/fulltext/72165.text.html
Result
N m
-2

Z / A Mean MW
incl. electrons
Average
Density
Sirius B Core Gas Pressure 4.78E+20 ? 1.6 [14/9]
uncertain estimate
3.00E+09
Sirius B Core Rad. P 2.04E+14
Sirius B Degeneracy P 3.11E+22 0.5

If these rough calculations are in the right ball park (and I am somewhat unclear about the average molecular
weight term), the Sun is shown to have high gas pressure, negligible radiation pressure compared to its gas
pressure, and small degeneracy pressure compared to its gas pressure. In contrast, Sirius B has very high
degeneracy pressure (much higher than the Suns core gas pressure), modest gas pressure compared to its
degeneracy pressure (though much higher than the Suns gas pressure), and negligible radiation pressure
compared to degeneracy gas pressure (but still higher than the Suns radiation pressure).

The Chandrasekhar Limit for White Dwarfs
Subrahmanyan Chandrasekhar (19101995) calculated in 1930 an upper size limit for a white dwarf, beyond
which the core is not adequately supported by its electron degeneracy pressure, resulting in a collapse to a
much denser neutron star.

/
[(
]
/

The radius Rwd of a white dwarf is thus inversely proportional to Mwd
1/3
, and in the non-relativistic
approximation, Mwd * Vwd = constant. Thus, volume of a white dwarf is approximately inversely proportional
to its mass M. For large enough mass,
relativistic considerations apply, dynamical
instability results (i.e., there is no stable
solution), the volume shrinks drastically, and
the white dwarf collapses to become a neutron
star. (This is related to the collapse of a
depleted stellar core lacking adequate total
core pressure, causing a core-collapse
supernova.) IMA2 p. 571 gives an approximate
formula for this MCh, but its underestimates
the currently accepted Chandrasekhar Limit of
Mch = 1.44 M.
The graph to the right
340
show the relationship
of radius to mass of white dwarfs. The lower
curve depicts the relativistic case, for which
mass falls to near-zero radius at M = MCh. This
is similar to the graph at IMA2 p. 571. I have
not determined the significance of the ultra-
relativistic limit.
The pressure P in a degenerate core is
independent of T, and as noted earlier, a runaway thermonuclear reactions can transiently occur at the
helium core flash of an evolving star at the Red Giant tip of the RGB phase.

340
http://en.wikipedia.org/wiki/Chandrasekhar_limit
Cooling of White Dwarfs
In a WD, degenerate electrons can travel long distances (because most of the electron states along their way
are already occupied). Electron conduction rather than photon radiation therefore is the principle means of
transporting heat toward the surface in the highly degenerate core, and its efficiency causes the interior of a
WD to be nearly isothermal. The outer thin non-degenerate envelope of the WD is the location where T
suddenly begins to fall and the T gradient leads to superficial convection. The surface of a WD cools more
slowly than its isothermal interior as the stars thermal energy leaks into space. (IMA2 p. 574, not absolutely
clear to me)
Most of the energy that be given up from cooling is in the nuclei, as the electrons are already mostly in their
lowest energy states and can go no lower. A rough calculation of available energy to be radiated away is
6x10
40
J, and this would be lost by luminosity at a cooling timescale of more than about 170 Ma. A cooling
curve for a 0.6 M WD (IMA2 p. 576) shows log (L/L) dropping from 1 initially to about -4 (1/10,000) in
about 5 Ga and to -5 (1/100,000) in about 15 Ga.
A fascinating aspect of WD cooling is that it will eventually undergo crystallization from the inside out. This is
the C-O core, which forms body-centered cubic lattice (BCC) crystals. The release of latent heat of
crystallization temporarily slows down the cooling curve at the knee of the curve... One of the more
memorable quotes in the textbook states, Thus the ultimate monument to the lives of most stars will be a
diamond in the sky, a cold, dark, Earth-size sphere of crystallized carbon and oxygen floating through the
depths of space. (IMA2 p. 576) Incidentally, the packing factor for a BCC crystal is 0.680,
341
tighter packing
than that seen with terrestrial diamonds (which have a diamond cubic crystal structure with packing factor of
only ~0.34).
342

Cooling of a pulsating WD can be
observed in the form of slowing of
pulsations. It is inferred from the
number density of WDs of various
ages (graph to right, vertical axis)
343

that the number of WDs falls off
rapidly at log (L/L) = 4.5. This leads
to a conclusion that WDs in the MW
galactic disk first began to form about
9 Ga ago, and that their progenitor
disk stars began to form at about 9.3
Ga ago (based on a mean pre-white-
dwarf lifetime of 0.3 Gyr). (However,
MW globular cluster stars have been
shown to be up to 3 Ga older.)
(IMA2 p. 578)

341
http://en.wikipedia.org/wiki/Cubic_crystal_system
342
http://en.wikipedia.org/wiki/Diamond_cubic
343
Winget DE, et al, An independent method for determining the age of the universe, Astrophysical Journal,
Part 2 - Letters vol. 315, April 15, 1987, p. L77-L81. http://adsabs.harvard.edu/doi/10.1086/184864
Neutron Stars (NS)
The neutron was discovered by James Chadwick (18911974) in 1932 and announced in a letter to the
editor.
344
Neutron stars were first proposed in 1934 by Walter Baade (18931960) and Fritz Zwicky (1898
1974), also as a letter to the editor.
345
They suggested that super-novae represent a transition from ordinary
stars to neutron stars.
Neutron stars are stellar remnants that form when increasingly degenerate stellar cores approach a mass
near or exceeding the Chandrasekhar Limit (~1.4 M). The core collapses, causing a core-collapse supernova
(Type Ib, Ic, or II, see Chapter 15 discussion) and leaves a neutron star stellar remnant. They resemble a
single confluent nucleus consisting mainly of neutrons, about 10
57
neutrons total, held together by gravity
which is opposed by neutron degeneracy pressure. Neutrons are also fermions subject to Pauli Exclusion
principle.
The radius of a neutron star is given by
/
[
]
/

This formula, which assumed a constant density, underestimates the radius somewhat. For a 1.4 M NS, the
formula gives a radius of 4.4 km, but the best estimates are 10 to 15 km.
346
For a NS of radius 10 km and
mass of 1.4 M, the average density would be 6.65x10
17
kg m
-3
. (At the center, the density is estimated at
10
18
kg m
-3
, see below.) By comparison, an atomic nucleus has a significantly smaller density of about
2.3x10
17
kg m
-3
(IMA2 p. 578 and here
347
), and the density of a nucleon is estimated at 3.0 x10
17
kg m
-3

(assuming a neutron has a classical radius of 1.1 fm and a mass of 1.67 x10
17
kg). Therefore, the neutrons
are tightly packed together, perhaps even overlapping.
The gravitational force at the surface has typical values of a few 10
12
m s
-2
up to a maximum of 710
12
m s
-2

(which is more than 10
11
times of that of Earth, 9.78 m s
-2
).
348
An object dropped from 1 m would accelerate
to an impact velocity of 1.9x10
6
m s
-1
. Calculations should use general relativity for most accurate results.
See Chapter 18 summary (and IMA2 p. 661) for estimation of mass release by matter infalling to a NS.

The Equation of State
The following discusses the sequence whereby an iron white dwarf at the center of a massive supergiant star
progresses to a neutron star. At the high pressure and density 10
9
kg m
-3
, such as seen in such an iron core
(IMA2 p. 579), the electrons become relativistic and interact with protons (electron capture) to create
neutrons:
p
+
+ e

n + e
The relativistic electrons supply the energy needed for the greater mass of the products, which differs from
that of the reactants by about 0.78 MeV. The density at which this reaction begins to occur is thought to be
1.2 x 10
10
kg m
-3
(IMA2 p. 580). Note that this value is somewhat higher than the estimated average white
dwarf density. (I have not yet found good estimates of core white dwarf density.) A more detailed calculation
determines that the actual density needed in an iron core for
56
Fe protons to capture relativistic free electrons
is even higher, >10
12
kg m
-3
. This process of neutronization reduces the Coulomb repulsion as a result of
declining proton packing density, and leads to nucleosynthesis of a series of nuclides:
56
Fe,
62
Ni,
64
Ni,
66
Ni,
86
Kr, ...,
118
Kr.(IMA2 p. 580 and here
349
). The neutron-rich nuclei do not decay by beta-decay, because there

344
Chadwick J, Possible Existence of a Neutron, Nature 129:312, 1932. doi:10.1038/129312a0
345
Baade W and Zwicky F, Remarks on Super-Novae and Cosmic Rays, Phys. Rev. 46, 7677 (1934)
346
See discussion of radiation radius given in Lattimer JM and Prakash M, Neutron Star Structure and the
Equation of State, The Astrophysical Journal, 550:426-442, March 20, 2001, http://arxiv.org/abs/astro-
ph/0002232
347
Nuclear Density: for
12
Carbon = (1.99264824x10
-26
) / ((4/3)**((1.2*(12
(1/3)
)*10
-15
))
3
) = 2.29x10
+17
kg m
-3
,
see http://www.chemistryexplained.com/Ar-Bo/Atomic-Nucleus.html
348
IMA2 p. 578 and http://en.wikipedia.org/wiki/Neutron_star
349
Wallerstein G et al, Synthesis of the elements in stars: forty years of progress, Rev. Mod. Phys. 69, 995
1084 (1997) .
are no vacant electron states available to take the electron. At 4x10
14
kg m
-3
, the core is a lattice mixture of
neutron-rich nuclei, free neutrons (neutron drip), and free degenerate electrons. The free neutrons pair up to
form bosons which are not subject to Pauli Exclusion Principle and therefore can all crowd into a superfluid
state of lowest energy. At a density of 4x10
15
kg m
-3
, neutron degeneracy pressure exceeds electron
degeneracy pressure. By density nuc 2x10
17
kg m
-3
, there is no longer a meaning distinction between inside
versus outside nuclei, and nuclei dissolve. The charged protons constitute a superconducting fluid. The
core has therefore become a fluid mixture of superfluid free neutrons, superfluid superconducting free
protons, and relativistic free electrons, in number ratios of 8:1:1. Pions begin to form when density is more
than twice nuclear density, as a result of decay of neutrons:
n

p
+
+

Modeling becomes increasingly difficult.
A recent calculation for a 1.4 M neutron star with radius 10.6 km (IMA2 p. 582-3, no reference given, see
also here
350
) suggests that there are the following layers:
Outer Crust, 10.3 to 10.6 km:
Consists of heavy nuclei (ions), nonrelativistic and relativistic degenerate electrons, a liquid or solid
lattice form, mostly
56
Fe near the surface, increasing neutron-rich nuclei with depth. The bottom is
the site of neutron drip at =4x10
14
kg m
-3
.
Inner Crust, 9.7 to 10.3 km:
Consists of somewhat heavier nuclei such as
118
Kr, a superfluid of free neutrons, and relativistic
degenerate electrons. The bottom is the site where nuclei dissolve at =nuc 2x10
17
kg m
-3
.
Interior (or Outer Core), 0 to 9.7 km:
A fluid mixture of superfluid free neutrons, a smaller number of superfluid superconducting free
protons, and a small % of relativistic free electron gas.
Core? (or Inner Core)
There may or may not be a solid inner core. The central density of a NS of this mass and size is
estimated to be 10
18
kg m
-3
.

Chandrasekhar Limit for Neutron Stars
The relationship of radius to mass in a roughly inverse
one, similar to white dwarfs, so they shrink and become
more dense with increasing mass. The graphs
351
to the
right illustrate the mass-radius relationship as computed
by various equations of state (EOSs) for stars
containing nucleons and, in some cases, hyperons. The
EOSs will not be further described here. Clearly,
however, radius shrinks as mass rises, but in a complex
manner.
Just as a WD collapses to a NS when electron degeneracy
can no longer support it, a NS has a maximal mass limit
(the Chandrasekhar Limit for Neutron Stars) beyond
which neutron degeneracy pressure can no longer
support it. This mass limit is 2.2 M if the star is not
rotating (unlikely) and 2.9 M if the star is rotating
rapidly (as a result of centrifugal force effects). When the
star exceeds these limits and becomes dynamically
unstable, it will collapse to a black hole.
Although IMA2 p. 583 uses the term The Chandrasekhar
Limit for Neutron Stars, this limit is not universally called
this name. Often it is unnamed, or named according to
the models employedincluding the original Tolman

350
http://en.wikipedia.org/wiki/Neutron_star
351
Lattimer JM and Prakash M, Neutron Star Structure and the Equation of State, The Astrophysical
Journal, 550:426-442, March 20, 2001, http://arxiv.org/abs/astro-ph/0002232
OppenheimerVolkoff limit.
352
The equations of state for neutron stars are evolving and newer models will
continue to appear.

Rotation and Angular Momentum
The slowly rotating stellar core becomes a rapidly rotating NS. The period of the NS is estimated as follows
(based on conserved angular momentum and the change in moment of inertia):
PNS 3.8 x 10
-6
Pcore.
It is difficult to estimate the rate of rotation of a progenitor stellar core, but IMA2 p. 584 uses a value of
1350 s as representative. This leads to a NS period of 5x10
-3
s (5 msec).
Neutron stars slow in rotation at a typical slow down rate of 10
15
seconds per rotation.
353

Magnetic Field
Neutron stars are expected to have intense magnetic fields because the field lines of a collapsing stellar core
are frozen in. The field strength therefore increases markedly compared to the stellar cores from which NSs
form. For spherically symmetrical objects, the magnetic field strength will be inversely proportional to the
surface area. With marked contraction, the field strength from the frozen-in field lines is greatly increased. If
an iron stellar core were to have a field strength of B 5x10
4
T (a maximum observed for free WDs, admittedly
not the same starting point), the resulting NS field strength is 1.3x10
10
T. The maximum B observed is 10
11
T
(10
15
G) in what are called magnetars.
354
Typical NS field strength B = 10
8
K or less. A formulas for B is given
below.

Neutron Star Temperatures and Emissions
Neutron stars begin very hot from their origin in a supernova. Their initial temperature of ~10
11
K rapidly
declines by a process that emits profuse neutrino radiation, the Urca process (named in 1941 after a casino in
Rio de Janeiro that rapidly depleted ones money, the Cassino da Urca). The neutrinos are formed in great
numbers and carry away much energy. In one day, the T drops to about 10
9
K. Interior T is 10
8
K after a few
hundred years, at which time the surface T has dropped to several million K. Surface T will remain at ~10
6
K
for ~10,000 years. At this Teff, the StefanBoltzmann law predicts a luminosity of 7.13x10
25
W, about 1/5 of
solar luminosity (L = of 3.846 x 10
26
W). The Planck distribution peak by Wien Displacement occurs at a
wavelength of 2.9 nm, corresponding to a photon energy of 0.4 keV, thus predominantly soft X-ray emissions.
This range is best studied by space-based X-ray observatories such as the Chandra X-ray Observatory (1999
), ROSAT (Rntgensatellit, 19901998), ASCA (19932001) and XMM-Newton (1999), etc.

Miscellaneous
Gravitational light deflection arises at a neutron star. Due to this GR light deflection, substantially more
than half of the spherical surface is visible, an effect that increases with increasing mass.
355

The weak visible light emitted is presumably perceived, if at all, as white or bluishwhite ( in view of the peak
emissions in soft X-ray).
There are about 2000 known neutron stars in the MW and Magellanic Clouds (per Wikipedia).

352
http://en.wikipedia.org/wiki/Tolman%E2%80%93Oppenheimer%E2%80%93Volkoff_limit
353
http://en.wikipedia.org/wiki/Neutron_star
354
http://www.cv.nrao.edu/course/astr534/Pulsars.html
355
http://www.spacetimetravel.org/galerie/galerie.html
Pulsars
The word Pulsar is a contraction of pulsating star. The pulsations of rotating neutron stars with strong
magnetic fields were first detected on November 28, 1967 by Jocelyn Bell Burnell (1943) and Antony Hewish
(1924). It was a signal that recurred every sidereal day (now known to be arising from pulsar PSR 1919+21,
where PSR=Pulsating Source of Radio). The pulse period P was 1.337 s (0.75 revolutions s
-1
or Hz). Although
PSR 1919+21 emits in radio wavelengths, pulsars have subsequently been found that emit in infrared, visible
light, X-ray, and/or gamma ray wavelengths.
356
Geminga (PSR J0633+1746) only radiates pulses in soft X-
rays and gamma rays (with period 237 ms)it is radio quiet. It has been identified in visible light, but I am
unclear whether there are visible light pulsations. (IMA2 p. 590 and here
357
)

Most known pulsars are in the MW galaxy or in nearby galaxies. There were about 1500 known pulsars as of
2005 (IMA2 p. 586, updated to 2008 in May 2012 according to ATNF
358
and 2123 according to SIMBAD.
359

They have periods mostly between 0.25 s and 2 s, with average pulsation period of 0.795 s (1.26 Hz). The
longest known pulsar period is 9.437 s (0.106 Hz)
360
or PSR 1841-0456 with period=11.8 s and rotation at
0.085 Hz (IMA2 p. 588 and here,
361
showing this star is subject to glitches).

Pulsar PSR J1748-2446ad (Terzan 5ad), discovered in 2005, is, as of 2010, the swiftest spinning pulsar
currently known, spinning at 716 times a second (716 Hz), therefore with a period=1.4 msec.
362

An X-ray pulsar, IGR J00291+593, spins at 599 s
-1
(599 Hz), thus with a period of 1.7 msec.

Pulsars, like other neutron stars, usually have a highly regular rate of rotation. The period normally
increases at a rate of about

,
(a unitless result from s
-1
/s
-1
). If
were to remain constant, the characteristic lifetime would be the time

required to bring the rotation to a halt, and calculated values range from a few tens of millions of years to
about 500 million years.

Pulsars as Rapidly Rotating Neutron Stars
Thomas Gold (19202004) first concluded that these pulsations must be due to rapid rotation of a neutron
star.

Possible mechanisms considered for pulsating stars with this short period include Binary Stars (inconsistent
with WDs and with binary NSs due to General Relativity), Pulsating or Oscillating WDs or NSs (infeasible
mathematically), and Rotating Stars (the favored explanation).

There is substantial evidence supporting an origin of NS from core-collapse supernovae. The NS often have a
high peculiar velocity, and few are found in binary systems. This may indicate that an asymmetrical
explosion or a jet propelled the NS from its starting point. By running the clock backwards, it appears the
Crab Nebula has accelerated in its expansion since its inception.

If a NS rotates too fast, it will be torn apart. The minimum rotation period a NS must have for which gravity
still overcomes centrifugal effects is given by (IMA2 p. 590):

356
http://en.wikipedia.org/wiki/Pulsar
357
http://en.wikipedia.org/wiki/Geminga
358
http://www.atnf.csiro.au/research/pulsar/psrcat/ click on Table without filling in any blanks
359
Use query criteria: otype = psr
360
Young MD et al, A radio pulsar with an 8.5-second period that challenges emission models Nature 400,
848-849 (26 August 1999) doi:10.1038/23650. This describes PSR J2144-3933
361
http://iopscience.iop.org/0004-637X/673/2/1044/fulltext/
362
Hessels JWT, et al, A Radio Pulsar Spinning at 716 Hz, Science 31 March 2006, Vol. 311 no. 5769 pp.
1901-1904, DOI: 10.1126/science.1123430

For a 1.4 M NS, the minimum allowed period is 5x10
-4
s = 0.5 msec. All pulsars to date have had periods
exceeding this lower limiting value and are therefore not in immediate risk of flying apart.

Many pulsars have been associated with supernova remnants, including the Crab Pulsar (PSR B0531+21 aka
PSR J0534+2200, rotation period ~33 msec) and the Vela Pulsar (PSR J0835-4510 aka PSR B0833-45, radio
period ~89 msec). The Crab Pulsars rotation rate is incompatible with a WD (as is Velas). (IMA2 p. 590)

Despite the precise measurements that have been made of rotation periods, pulsars do not always maintain a
steady and predictable increase in period. (For this reason, I believe that they will never make good
astronomical time-keeping devices.) Young pulsars, including the Vela and Crab pulsars, are subject to
glitches (see below).

Pulsation of Crab Pulsar: A time sequence for the pulsar in the Crab nebula is shown, and its
location in the Crab nebula (as shown in a KPNO 4-meter Mayall image [1973]). The images of
this sequence were also obtained with the 4-meter Mayall telescope on Kitt Peak, during the
night of 20 October 1989 using a standard B-band optical filter and the Kitt Peak Photon
Counting Array (KPCA). Phased accumulation over almost 2 hours was necessary to create
this image; the observed period that day was 33.36702 milliseconds. Each of the 33 images
represents a time slice of about 1 millisecond in the pulsar period. The brighter, primary pulse
is visible in the first column: the weaker, broader inter-pulse can be seen in the second
column. (See graph next page of pulse sequences.) Sequence of visible light images are
arranged from top to bottom, then left to right. Foreground star is invariant while Crab Pulsar
varies in light output. (See light curve.) (N. A. Sharp / AURA / NOAO / NSF)
363

363
http://messier.seds.org/more/m001_pulsar.html
The following image also depicts the Crab Pulsar in the center of the Crab Nebula, but it is much higher in
resolution (from HST). It shows a knot of hot emission very near the pulsar (slightly southeast of the pulsar,
separated from it by only 1200 AU). It also shows that in the direction opposite the knot, the Crab pulsar is
capped by a ring-like halo' of emission tipped at about 20 degrees to our line of sight. In this geometry the
polar jet flows right through the center of the halo. The newly discovered ring may mark the boundary
between the polar wind and jet, and an equatorial wind that powers a larger torus of emission surrounding
the pulsar. (HST, Hester and Scowen, 1995):
364

Crab Pulsar Light Curve: The period of the Crab Pulsar
pulsations is measured between the higher peaks. One full
revolution of the neutron star corresponds to the distance
from one high peak to the next, and the diagram therefore
covers six consecutive revolutions (about 200
milliseconds).
365
The weaker, broader inter-pulse peaks are
also evident (image to right)
The Crab Pulsar period has varied from 33.35 ms in May
1988 to 33.67 ms in May 2012,
366
with a few glitches along
the way. (Period in msec is given by 1000/(Nu Hz.)

364
Hester et.al., WFPC2 Studies of the Crab Nebula. I. HST and ROSAT Imaging of the Synchrotron Nebula.
Astrophysical Journal, Vol. 448, p. 240-263 and plate 10-15 (July 20, 1995) Image obtained from
http://messier.seds.org/more/m001_hst.html#hester1995 Also found at IMA2 p. 594.
365
http://www.eso.org/public/images/eso9948i/ including image of light curve
366
http://www.jb.man.ac.uk/pulsar/crab/crab2.txt
Synchrotron Radiation and Curvature Radiation
The Crab Nebula is connected with its central pulsar. The ghostly glow the nebula exhibits was proposed (by
Iosif Shklovsky, 1916-1985)) to be synchrotron radiation. Synchrotron radiation is generated by acceleration
of relativistic electrons as a result of the Lorentz force F (where F = q (E + v B) exerted in the presence of a
magnetic field B. The resulting acceleration causes them to circle or spiral around often curving magnetic
field lines. E is negligible here. The acceleration of charged particles in a magnetic field leads to emission of
synch. rad according to the right hand rule (where v=index finger, B=middle finger, F=thumb). (A related
radiation, curvature radiation, arises from electrons that are moving along the curving field lines.) The non-
thermal emission is a strongly linearly polarized but otherwise featureless broad spectrum continuum with
energy distribution which depends on the velocity distribution of the electrons, and which is quite distinct
from blackbody (aka thermal) radiation. For relativistic electron velocities, the emission pattern is sharply
collimated forward (relativistic beaming) in the direction of the electrons current velocity vector. Parts of the
Crab Nebula emit light which is 60% linearly polarized, confirming its origin as synchrotron radiation.
The relativistic electrons and the synchrotron radiation should have long ago died away in the Crab Nebula if
they were simply residuals from the SN. Instead, it appears that the central pulsar is replenishing both the
relativistic electrons and the magnetic field. The calculated amount of energy being injected is a remarkable
5x10
31
W or 10
5
L. The NS is acting as a huge flywheel with a high amount of rotational kinetic energy,
energy gradually imparted to the surrounding nebula. Using previously stated average values for a 1.4 M
NS, the expected kinetic rotational energy loss resulting from the documented slowing of the pulsar matches
well the energy needed to replenish the nebula. The required energy is not supplied by the radio pulses in the
Crabthey are much too weak, only about 10
24
W. The wisps of the Crab Nebula are moving outward at
velocities as high as 0.5c. (IMA2 p. 595)

Pulse Characteristics
The radio pulses are usually received at 20 MHz
to 10 GHz. Passage of radio waves through the
ISM (or nebular media) causes the ISM
electrons to oscillate, the net effect being that
the waves are retarded somewhat below
vacuum c, with greater retardation at lower
frequencies. This causes a spreading out or
dispersion of the pulses. (See diagram to right,
which shows Uncorrected dispersive delays for
a pulsar observation over a bandwidth of 288
MHz (96 channels of 3 MHz width each),
centered at 1380 MHz. The delays wrap since
the data are folded (i.e. averaged) modulo the
pulse period.).
367
Note that this depicts
dispersion by frequency, not waveform drift
with time or subpulses. The amount of
dispersion may be used to estimate distance to
the pulsar (at best to ~30% d/d accuracy).
Since pulsar observations almost always cover
a wide bandwidth, uncorrected differential
delays across the band will cause dispersive
smearing of the pulse profile. For pulsar
searches, the DM is unknown and becomes a
search parameter much like the pulsar spin
frequency. This extra search dimension is one

367
Pulse Dispersion:
all quotes from http://www.cv.nrao.edu/course/astr534/Pulsars.html,
diagram from http://www.jb.man.ac.uk/research/pulsar/handbook/figures/med/1.08.gif
see also http://astronomy.swin.edu.au/cms/astro/cosmos/p/Pulsar+Dispersion+Measure
of the primary reasons that pulsar searches are computationally intensive.
Despite the Crab Nebula light curve, pulses in general are very brief and occupy only a small part of the
period, typically 1 to 5% in the 20 MHz to 10 GHz band. The individual pulses vary somewhat, but the
averaged waveform from integrating 100 pulses tends to be stable. Pulsars may have more than one wave
pulse mode and switch between these pulse modes. There may be subpulses, including ones that drift, and
the different modes may have different subpulses, up to 5 or more. (IMA2 p. 598). Pulses may disappear
(null) for 100 periods or so, then return. They are strongly
linearly polarized.

The graph to the right shows a pulse-stack of fifty
successive pulses of PSR B0818-13. Two successive drift
bands are vertically separated by P3 [measured in periods]
and horizontally by P2 [measured in phase change].
368

Pulsars may be grouped with a graph of period
versus period time derivative, the so-called

diagram (see graph on left).

369
This is
useful for following the lives of pulsars, playing
a role similar to the Hertzsprung-Russell
diagram for ordinary stars. It encodes a
tremendous amount of information about the
pulsar population and its properties, as
determined and estimated from two of the
primary observables, P and Pdot. Using those
parameters, one can estimate the pulsar age,
magnetic field strength B, and spin-down
power
.
370
Here, SNR=supernova remnant,
SGR=Soft Gamma Repeaters, AXP=Anomalous
X-ray pulsars.

368
Weltevrede, P., The subpulse modulation properties of pulsars at 92 cm and the frequency dependence of
subpulse modulation Astronomy and Astrophysics, Volume 469, Issue 2, July II 2007, pp.607-63, DOI:
10.1051/0004-6361:20066855
369
http://www.jb.man.ac.uk/research/pulsar/handbook/figures/med/1.13.gif
370
www.cv.nrao.edu/course/astr534/Pulsars.html (Good review of pulsar properties)
The following rather complex colored graph depicts stacked pulse graphs for Pulsar PSR B0809+74 at 328
MHz, and explores the distribution of polarization:
371

Color display of a 200-pulse portion... The first column of the displays [on the left] gives the
total intensity (Stokes parameter I [total power]), with the vertical axis representing the pulse
number and the horizontal axis pulse longitude [relative phase], colour-coded according to the
... top bar to the left of the displays. The second and third columns give the corresponding
fractional linear polarisation L/I ... and its angle ... The last column gives the fractional
circular polarisation (V/I)... Note the variation in pulse peak total power, and the drift in
pulse phase. The drifting subpulses ... exhibit strikingly linear and circular polarization
which appears to reflect the characteristics of two nearly orthogonally polarized
emission modes.

371
Ramachandran R et al, Pulsar drifting-subpulse polarization: No evidence for systematic polarization-
angle rotations, A&A 381, 993{999 (2002), DOI: 10.1051/0004-6361:20011525
Pulsar Model and Postulated Emissive Mechanism
The traditional model is of a rotating
neutron star in which the axis of rotation
differs significantly from the magnetic
dipole field axis:

Diagram
372
to right depicts the traditional
magnetic dipole model of a pulsar and its
generation of the radio beam (emission
cone).

See discussion below.

The emissive mechanism is poorly
understood, but a model is described as
follows (from here
373
and with my
interpolations in [] )

The radio pulses originate in the pulsar
magnetosphere. Because the neutron star
is a spinning magnetic dipole, it acts as a
unipolar generator. The total Lorentz force
acting on a charged particle is [for SI units,
F = q (E + v B)].

Charges in the magnetic equatorial region
redistribute themselves by moving along
closed field lines until they build up an
electrostatic field large enough to cancel the magnetic force and give |F|=0 . The voltage induced is about
10
16
V in MKS units. However, the co-rotating field lines emerging from the polar caps cross the light
cylinder (the cylinder centered on the pulsar and aligned with the rotation axis at whose radius the co-
rotating speed equals the speed of light) and these field lines cannot close. Electrons in the polar cap are
magnetically accelerated to very high energies along the open but curved field lines, where the acceleration
resulting from the curvature causes them to emit curvature radiation that is strongly polarized in the plane of
curvature. As the radio beam [of curvature radiation] sweeps across the line-of-sight, the plane of polarization
is observed to rotate by up to 180 degrees, a purely geometrical effect. High-energy photons produced by
curvature radiation interact with the magnetic field and lower-energy photons to produce electron-positron
pairs that radiate more high-
energy photons. The final
results of this cascade process
are bunches of charged
particles that emit at radio
wavelengths. (Some of these
steps are depicted in the
diagram to the left].
374

The death line in the

diagram [beyond which lies the
graveyard] corresponds to
neutron stars with sufficiently
low B and high P that the
curvature radiation near the
polar surface is no longer
capable of generating particle

372
http://www.jb.man.ac.uk/research/pulsar/handbook/ch3.html
373
www.cv.nrao.edu/course/astr534/Pulsars.html All quotes in this section are from this NRAO webpage
except as noted.
374
http://www.jb.man.ac.uk/research/pulsar/handbook/figures/med/3.03.gif
cascades. [In other words, at this point radio pulsations cease.]

The extremely high brightness temperatures are explained by coherent radiation. The electrons do not
radiate as independent charges e; instead bunches of N electrons in volumes whose dimensions are less than
a wavelength emit in phase as charges Ne . Since Larmor's formula [for radiation from a rotating electric
dipole] indicates that the power radiated by a charge q is proportional to q
2
, the radiation intensity can be N
times brighter than incoherent radiation from the same total number N of electrons. Because the coherent
volume is smaller at shorter wavelengths, most pulsars have extremely steep radio spectra. Typical (negative)
pulsar spectral indices are 1.7 (S
-1.7
), although some can be much steeper ( > 3) and a handful are
almost flat ( < 0.5).

The energy (J s-1) emitted due to a rotating magnetic dipole is given by (IMA2 p. 597):

The magnetic field at the pole of a NS is given by (IMA2 p. 599):

where I is the moment of inertia, R = star radius, = angle between rotation axis and magnetic axis.

The radius of the light cylinder is given by Rc = c/ = cP/2.

The strong electric field continuously rips electrons and positive ions from the surface, easily overcoming
gravity, causing the rotating magnetosphere to be filled with charged particles. At the light cylinder, the
particles cannot keep up with the co-rotation (which is at the speed of light) and are spun away as part of a
pulsar wind that carries away part of the magnetic field as well. This kind of wind may explain the
mechanism for replenishment of the Crab Nebulas energy. (IMA2 p. 601)

According to the model described in IMA2 p. 601, which differs only a little from the NRAO description,
electrons and ions in the polar region are accelerated to relativistic speeds by the induced electric field. As the
electrons follow the curved magnetic field lines, they emit curvature radiation that is relativistically formed
into a narrow beam. Gamma ray photons produced by curvature radiation can spontaneously produce
electron-positron pairs that subsequently are accelerated again radiate more high-energy photons, and the
cascading process continues. Coherent beams of curvature radiation emitted by bunches of particles may be
responsible for the individual subpulses. The narrowly directed radio wave beams emitted near the polar
region sweep around the star light a lighthouse beacon might. A terrestrial detector detects the pulses as why
rotate past the Earth.

Glitches are very sudden though infrequent changes in pulsar period. One seen for Vela in 2010 exhibited a
relative change in frequency / of 1.94x10
-6
.
375
Glitches may be a result of starquakes resulting from
fractures, faulting, and shifting in the rigid solid crust, or perhaps an unpinning of superfluid vortices in the
core of the NS, but this is a subject requiring more research. (IMA2 p. 602)

Fate of a Pulsar
It is unclear whether the magnetic field must decay with time. In one view, this decay would take place with a
characteristic time of perhaps 9 million years, which will lead eventually to cessation of detectable pulses.
Alternatively, the magnetic field may not decay. But in either case, the rotation rate will slowly decrease and
this will cause the beam to become weaker even if the magnetic field does not changes. In the
diagram,
they will enter the graveyard and become a radio-quiet NS. (IMA2 p. 602)

375
http://www.hartrao.ac.za/news/100731vela/index.html See also IMA2 p. 591.
Soft Gamma Repeaters
It is possible that magnetars are responsible for soft gamma repeaters SGRs. These objects emit hard X-rays
and soft gamma rays up to 100 keV. Only a few are known, and these mostly correlate with SNRs. One
detected in August 2008 by a giant X-ray outburst, is SGR 0501+4516. It is estimated to lie about 15 000
light-years away, and was undiscovered until its outburst gave it away
376
The sudden energy outburst of an
SGR is thought to be caused by a crack in the crust that allows a release of energy greatly exceeding the
Eddington luminosity limit for stars, though it is very short-lived. (IMA2 p. 603) SGR 0501+4516 was
observed with XMMNewton and Swift-XRT quasicontinuous monitoring and its pulses were fitted to
blackbody radiation and power law curves. The data strongly argue that this source is a magnetar candidate
with a magnetic field of B ~210
10
T. The discovery of SGR0501+4516 , and its AXP-like characteristics,
represents another piece of evidence in the unification of the magnetar candidate class, weakening further the
differences between AXPs, TAXPs, and SGRs.
377
Here, AXPs = Anomalous X-ray Pulsars, and TAXPs =
Transient Anomalous X-ray Pulsars. SGRs may arise in young SNRs (IMA2 p. 603) and may be relatively
short-lived. However, at least some AXPs are not associated with SNRs. At least one SGR has been described
for which the surface magnetic field is much lower than the usual range for magnetars, with B not greater
than 7.5 10
12
gauss (7.5 10
8
T)
378

376
Rea N, et al, The first outburst of the new magnetar candidate SGR 0501+4516,
http://arxiv.org/abs/0904.2413v1
377
ibid.
378
Rea N., et al, A Low-Magnetic-Field Soft Gamma Repeater, Science 12 November 2010, Vol. 330 no. 6006
pp. 944-946, DOI: 10.1126/science.1196088
General Relativity and Black Holes (Chapter 17, mostly omitted)
Spacetime tells matter how to move; matter tells spacetime how to curve.
379

This chapter has not yet been read in my astronomy courses and is not summarized here, aside from an
excerpt below.

Schwarzschild radius
The Schwarzschild radius is named for Karl Schwarzschild (18731916),
380
who derived this first exact
solution in the Einstein field equations of general relativity. It defines the event horizon of a black hole. For
a non-rotating star or black hole, it is given by

Black holes are fully described by the mass, angular momentum, and electric chargeexternally at least, a
black hole has no hair (IMA2 p. 634640). The maximum angular momentum for a black hole is
c

Rotating black holes are surrounded by a region of spacetime in which it is impossible to stand still, called
the ergosphere. This is the result of a process known as frame-dragging; general relativity predicts that any
rotating mass will tend to slightly "drag" along the spacetime immediately surrounding it. Any object near the
rotating mass will tend to start moving in the direction of rotation. For a rotating black hole this effect
becomes so strong near the event horizon that an object would have to move faster than the speed of light in
the opposite direction to just stand still... The ergosphere of a black hole is bounded by the (outer) event
horizon on the inside and an oblate spheroid, which coincides with the event horizon at the poles and is
noticeably wider around the equator. The outer boundary is sometimes called the ergosurface. Objects and
radiation can escape normally from the ergosphere. Through the Penrose process, objects can emerge from
the ergosphere with more energy than they entered. This energy is taken from the rotational energy of the
black hole causing it to slow down.
381

For a rotating black hole, the event horizon (like the ergosphere) is oblate and wider at the equator. For a
maximally rotating black hole, the event horizon at the equator is (IMA2 p. 641):

379
Wheeler JA, Ford KW, Geons, Black Holes, and Quantum Foam: A Life in Physics, p. 235, 1998.
380
http://en.wikipedia.org/wiki/Karl_Schwarzschild
381
http://en.wikipedia.org/wiki/Black_hole
Close Binary Star Systems (Chapter 18)

Gravity Potentials in a Close Binary Star System
Binary systems appear to be among the astrophysicists best friends. Binary systems are common, but in
most, the two components are widely spaced and negligibly interact. In a close binary system, the stars are
sufficiently close that at least one gravitationally distorts the others atmospheretheir separation is roughly
equal to the diameter of the larger star. A rotating star that is tidally deformed pulsates and dissipates energy
as heat. Such a system may eventually reach a state of minimum energy for its angular momentum,
resulting in circulazed orbits and, at least for normal sized non-compact stars, synchronous rotation (in
which, like the Earths moon, the same hemisphere always faces the same hemisphere of the other star). In
this situation, no further energy is lost to tidal interactions. (IMA2 p. 653 and here
382
)
Reference Frame: When referring to orbital rotation in a binary system, in all cases each body actually orbits
around the barycenter or center of gravity of the system. It is convenient to analyze binary system motion
using a co-rotating coordinate system with the same period as the two co-orbiting bodies and centered on the
barycenter (center of mass).

Equipotential Surfaces and Lagrangian Points
In the diagram to the right in a rotating frame of
reference, a Sun-like star is depicted at the center
orbited by an Earth-like planet having an orbiting
satellite moon. The actual barycenter is not
shown. The red arrowheads indicate downhill
gradients (i.e., the direction in which gravitational
potential energy U for a test mass decreases to a
more negative value) and the blue arrowheads
indicated uphill gradients. The force on a test
mass is given by
, and it is always
perpendicular to the equipotential contour.
The forces on a test object are strongest when the
contours of the effective potential are closest
together and weakest when the equipotential
contours are far apart.
The diagram depicts the 5 Lagrangian points (Lagrange points): The five positions in an orbital
configuration where a small object of negligible mass and affected only by gravity can theoretically be
stationary relative to two larger objects (such as a satellite with respect to the Earth and Moon). The
Lagrangian points mark positions where the combined gravitational pull of the two large masses provides
precisely the centripetal force required to cause the small object to orbit with the larger masses about the
barycenter.
383
L1, L2 and L3 are unstable, meaning that any small perturbations can dislodge an object at any
of these.
The effective gravitational potential energy per unit mass (of a test object) is given by (IMA2 p. 655):
(

At the precise L1, L2, and L3 points, there is no net force on a test mass because
(IMA2 p. 655).
However, a test mass at these points is unstable (some like ESA say metastable, others semi-stable) because

382
http://en.wikipedia.org/wiki/Binary_star
383
Lagrange (Lagrangian) Points:
http://map.gsfc.nasa.gov/mission/observatory_l2.html (from WMAP, which orbits about L2, as will JWST)
http://en.wikipedia.org/wiki/Lagrangian_point (color diagram shown)
http://map.gsfc.nasa.gov/media/ContentMedia/lagrange.pdf (mathematical derivation)
they represent local maxima of . More precisely, these points perch at the apex of a saddle, and almost all
perturbations cause the mass to accelerate downhill more or less in the direction of one of the blue
arrowheads.
The L1 point is most relevant in the study of close binary systems. L1 is also good for observing the Sunas
with SOHO, the Advanced Composition Explorer ACE, and the Global Geospace Science (GGS) WIND satellite.
Approximate formulas for the distance from L1 to M1 and L1 to M2 are given at IMA2 p. 656.
The L2 point is a good location for placing a satellite to observe the universe while protected by the Earth
from solar radiation (such as WMPA, JWST, Gaia, Herschel, and Planck).
The L3 point would be concealed by the Sun and is not suitable for satellites.
The L4 and L5 points (triangular or Trojan Lagrange points) are located about 60 degrees ahead of or behind
a secondary mass (like the Earth) orbiting a larger mass (like the Sun). The centres of mass of M and m and
the Lagrange point L5 form an equilateral triangle with the edge length r.,
384
where r is the length between
the centers of mass of M the primary mass and m the secondary mass. They are at the apex of a gravitational
potential hill falling off in all directions, and again with
. However, when perturbed, instead of

accelerating away as might be expected, a test mass moving away from an L4 or L5 point is acted on by the
Coriolis force, causing it to go into a stable circular orbit around the point. Therefore, an object at an L4 or L5
point is truly stable. They may someday be employed for space station locations. [Natural] Objects found
orbiting at the L4 and L5 points are often called Trojans after the three large asteroids Agamemnon, Achilles
and Hector that orbit in the L4 and L5 points of the Jupiter-Sun system... There are hundreds of Trojan
Asteroids in the solar system. Most orbit with Jupiter, but others orbit with Mars. In addition, several of
Saturn's moons have Trojan companions. No large asteroids have been found at the Trojan points of the
Earth-Moon or Earth-Sun systems. However, in 1956 the Polish astronomer Kordylewski discovered large
concentrations of dust at the Trojan points of the Earth-Moon system. Recently, the DIRBE instrument on
the COBE satellite confirmed earlier IRAS observations of a dust ring following the Earth's orbit around the
Sun. The existence of this ring is closely related to the Trojan points, but the story is complicated by the
effects of radiation pressure on the dust grains.
385
See NASA animation of the first known Earth Trojan, 2010
TK7.
386

Roche Lobe, Roche Limit, and Roche Sphere
The Roche lobe (named for douard Albert Roche, 1820 1883) is the approximately cylindrically
symmetrical 3D region of space around a star in a close binary system within which orbiting material is
gravitationally bound to that star. The Roche lobe for close binaries has an approximately tear-drop 3D
shape. If portions of the star expand past its Roche lobe (typically the atmosphere), then the atmospheric
material can escape the gravitational pull of that star. It may pass into the circumbinary region that
surrounds both stars, such as just inside the dumbbell shape that passes through L2.
387
But in a close binary
system the atmospheric material can most easily escape one Roche lobe by passing in to the companion stars
Roche lobe through the inner Lagrangian point L1. The apex of each Roche lobes tear-drop is at the L1
Lagrangian point of the system.. (See diagram that follows, which nicely depicts the relatively low potential
barrier that exists for passage of material through L1 compared to other nearby higher routes.) The Roche
lobe is different from the Roche limit (below) and the Roche sphere, which approximates the gravitational
sphere of influence of one astronomical body in the face of perturbations from another heavier body around
which it orbits.
388

The Roche limit (less relevant to this discussion) is the distance or radius within which a secondary or
satellite smaller celestial body M2, held together solely by its own self-gravity, will disintegrate due to the
distorting effect of the primary larger celestial body M1. The latter exerts tidal forces exceeding M2s
gravitational self-attraction. Tidal forces become destructive when portions of M2 orbit closer to the M1

384
http://esamultimedia.esa.int/docs/edu/HerschelPlanck/EN_13e_L_Points_EarthMoonSystem.pdf
385
http://map.gsfc.nasa.gov/mission/observatory_l2.html
386
http://www.nasa.gov/multimedia/videogallery/index.html?media_id=103550791
387
http://space.about.com/od/glossaries/g/rochelobe.htm
388
http://en.wikipedia.org/wiki/Hill_sphere
primary attempt to move more quickly (orbit at higher angular velocity) than portions farther away. (IMA2 p.
723 and here
389
)

3D Equipotential surface showing potential wells for two close binary stars;
also, a central plane cross section showing equipotential curves and lines,
the tear-shaped Roche lobes separated by L1,
the dumbbell passing through L2, and the L3 Lagrangian point.
390

Classes of Binary Star Systems
Close binaries are classified as detached (neither fills its Roche lobe), semi-detached (one fills its Roche
lobe), or contact (both fill their Roche lobes).
Detached binary stars are widely separated (radii are much smaller than their separation), maintain a
spherical shape, and have little influence on each other.
A semi-detached star in a binary system can expand beyond its Roche lobe and contribute atmospheric
material to its companion. The overflowing star is usually called the secondary and labeled M2. The
companion star which has not filled its Roche lobe and which receives the material from M2 is designated the
primary or M1.
Contact binaries share a common atmosphere bounded by a dumbbell-shaped equipotential surface and
freely exchange atmospheric material.

389
http://en.wikipedia.org/wiki/Roche_limit
390
http://hemel.waarnemen.com/Informatie/Sterren/hoofdstuk6.html#h6.2
Mass Transfer Rate and Release of Gravitational Energy
The rate of mass transfer
for a semi-detached binary with mass equaling its companion is roughly

estimated at (IMA2 p. 658):

where R is the radius of the expanded star, d is the distance over which of the two stars would overlap if they
both had the radius of the expanded star, and mH is the mass of a hydrogen atom. The more the stars
overlap, the greater is d, , and the T at the region of transfer, markedly increasing the transfer rate. Typical
mass transfer rates for semidetached binaries range from 10
-11
to 10
-7
M y
-1
. By comparison, solar wind
transports only about 3x10
-14
M y
-1
(IMA2 p. 660).
For a white dwarf star of mass 0.85 M and radius 0.0095 R, the potential gravitational energy U of an
infalling test mass that is converted to KE (and ultimately to light and heat) is 1.71 x 10
13
J kg
-1
, or 0.019%
of the rest energy (given by mc
2
) of the test mass.
However, for a neutron star of mass 1.4 M and radius 10 km = 1.4x10
-5
R, the potential gravitational
energy of an infalling test mass that is converted is 1.86 x 10
16
J kg
-1
, or 21% of the total mc
2
rest energy of
the test mass. This is a prodigious amount of energy, 30 times what would be released from hydrogen fusion
of the same mass of hydrogen. (IMA2 p. 661)
The observed luminosity of celestial X-ray sources are typically about 10
30
W. If this power arises from
infalling of mass to a neutron star, the mass transfer rate required is only about 10
-9
M yr
-1
.

Circumstellar Accretion Disks

Close binary system with red giant secondary star overflowing its Roche lobe across L1
into an accretion disk encircling a WD primary star.
The system is rotating counterclockwise, and this causes
the inflowing gas from the secondary to miss the white dwarf.
Instead, the rapid inflow first comes in contact with the periphery
of the accretion disk at a bright hot spot.
391

391
http://www.pas.rochester.edu/~afrank/A105/LectureXII/LectureXII.html
The orbital motion of a semidetached binary often prevents the direct infalling of the transported matter to the
primary (companion), usually when the primary has a radius of <5% of the binary separation a (measured
center to center). Instead the material goes into orbit about the primary to form a thin accretion disk in the
plane of the orbits shared by the two stars.
Accretion Disk Hot Spot
The first region of contact, where infalling material collides with the
accretion disk, can appear as a hot spot of enhanced intensity. This can
be seen in the light curves of eclipsing binary systems and reconstructed
into a circular intensity distribution. (Reconstructed negative image of
LX Ser to right, slightly contrast enhanced by MCM)
392

Accretion Disk Viscosity and Heating
The gases and particles in the accretion disk collide and slow from
viscosity (caused by still uncertain mechanisms). This process heats
the disk, with greatest T attained in the innermost disk region, and light
(or other EM) blackbody radiation is emitted. As the gas and particles
slow, they spiral inward and eventually reach the surface of the primary.
The migration period takes a few days to weeks. (IMA2 p. 666)

Accretion Disk Temperature, Luminosity, and Eddington Accretion Limit
The following formula best estimates the temperature of an accretion ring at a particular radius r
(IMA2 p. 663):
T (
)
/
(
)
/
( /)
/
T
)
/
( /)
/

where the characteristic temperature Tdisk is defined by
T
)
/

R and M are the radius and mass of the primary star,
is the rate of matter infalling, = the Stefan

Boltzmann constant, and r is a radius measured from the center of the primary.
The characteristic temperature Tdisk has a value approximately twice the maximum disk temperature (thus
it is an overestimate), which occurs at r 1.36 R. (IMA2 p. 707)
The luminosity of the accretion disk is given by (IMA2 p. 664):

The luminosity of the accretion disk shown here is 1/2 of the total luminosity arising from the infalling
matterthe other half comes from impact at the star surface (rather than at the accretion disk).
The maximal rate of mass transfer to the accretion disk at which the Eddington accretion limit is reached is
roughly given by
393

392
Rutten RGM, et al, Reconstruction of the accretion disk in six cataclysmic variable stars, Astronomy and
Astrophysics , vol. 260, no. 1-2, p. 213-226, 07/1992.
393
Philipp Podsiadlowski, http://www-astro.physics.ox.ac.uk/~podsi/lec_c1_4_c.pdf
or 1.8 x 10
-8
M y
-1
for a neutron star. Here, 0.034 m
2
kg
-1
is Rosseland mean opacity, and R is the
radius to the inner edge of the accretion disk. At this rate of mass deposition, the inner accretion disk is
raised to heights reaching 2R, where R is the radius of the NS (IMA2 p. 693).
For the same typical white dwarf WD used above, with mass 0.85 M and radius 0.0095 R, and for
=1.6x10
-10
M yr
-1
, the Tmax of the disk is 2.62x10
4
K, the peak is 111 nm (ultraviolet region), and the
luminosity Ldisk = 8.55x10
25
W = 0.22 L.
For the same typical neutron star NS used above, with mass 1.4 M and radius 10 km = 1.4x10
-5
R, and for
=1.6x10
-9
M yr
-1
, the Tmax of the disk is 6.86x10
6
K, the peak is 0.423 nm (X-ray region), and the luminosity
Ldisk > 2400 L! (IMA2 p. 665) Thus the accretion disks of neutron stars can be strong emitters of X-rays.
Accretion Disk Radial Extent
The accretion disk may not extend all the way to the surface of the primary star, due to magnetic field effects.
The outer limit of the accretion disk Rdisk is estimated in terms the radius rcirc at which matter infalling from
L1 first settles into a circular orbit.
[ o
( +
)
Here, a is again the binary separation (measured center to center).
For contact or semi-detached binaries, the transfer of mass from a star to its companion can also result in a
significant transfer of angular momentum. Some of the mass that crosses L1 into the primarys Roche lobe
will carry a higher than average angular momentum, causing it to migrate to the outer edge of the disk. There
it may be blown by the stellar wind, perhaps reaching the Roche lobe boundary and spilling into the
circumbinary region or being blown from the binary system by the stellar wind, carrying away some of the
systems angular momentum, just as the solar wind carries away angular momentum. (Gravitational
radiation is another mechanism for carrying away angular momentum from relativistic binary systems, and is
primarily responsible for loss of angular momentum in some short orbital period systems. (IMA2 p. 669 and
here
394
))
The gas that reaches the primary may add to the rotation of that star. Although tidally locked stars are not
rotating with respect to each other, some binary systems have more complex stellar rotations and rotation
axes. The text states however that stellar rotation angular momentum may be neglected. (IMA2 p. 668)

A Survey of Interacting Binary Systems
As binary systems evolve, the ratio of their stellar masses and individual angular momenta change.

Effect of Mass Transfer
As mass is transferred, the orbital separation changes, according to this formula (IMA2 p. 669):

Assuming as before that M1 > M2, mass transfer from the less massive star M2 to the more massive one M1
will make
positive, thus cause the RH to be positive, and the orbital separation a will increase. Mass
transfer from the more massive one M1 to the less massive star M2 will make
negative, thus cause the RH

to be negative, and the orbital separation will shrink. (Inferred from IMA2 p. 669, explicit here.
395
)
As the orbital separation decreases, the angular frequency increases (orbital period decreases), according to
(IMA2 p. 669):

394
http://arxiv.org/pdf/0811.0455
395
http://astronomy.nju.edu.cn/~lixd/XRB/chapter02.pdf

Evolution of a Binary System
There are numerous possible evolutionary pathways for close binary systems (see IMA2 p. 669673).
The following diagram (IMA2 p. 670-1 and here
396
) shows a representative sequence for a close binary system.
It begins with two well-separated main sequence stars of similar mass and evolves to two carbon-oxygen (C-O)
white dwarfs, one of which becomes a cataclysmic variable or Type Ia supernova. One star, M1 (on left),
becomes a red giant and overflows its Roche lobe to transfer mass to the less massive star M2. This causes
the stars to spiral closer together, producing a contact system with a common atmospheric envelope. In one
scenario, these two cores might merge to form a blue straggler. But, as depicted here, the envelope may be
ejected (similar to a planetary nebula). In this case, the system again emerges as a detached binary, with the
primary M1 now a white dwarf with exposed helium core, evolving to a C-O WD. Mass begins to transfer from
M2 to a new common envelope, and the stars spiral apart (per IMA2 p. 670) or come closer (per the article and
the diagram) as M2 evolves also to a C-O WD. The larger less massive WD M2 dissolves into a heavy disk that
is accreted by the remaining WD M1. Eventually, the remaining WD M1 is pushed over the Chandrasekhar
limit and explodes as a Type Ia supernova.

is the frequency (yr

-1
) in the galaxy of SNe Ia from 2 C-O
WDs by the close channel; is the duration of a stage; Porb is the orbital period; A is the semi-major axis.

396
Iben I, Tutukov AV, Supernovae of type I as end products of the evolution of binaries with components of
moderate initial mass (M not greater than about 9 solar masses), Astrophysical Journal Supplement Series,
vol. 54, Feb. 1984, p. 335-372, http://adsabs.harvard.edu/abs/1984ApJS...54..335

Types of Interacting Binary Systems
We define an active binary as a binary stellar system in which the activity level is significantly enhanced as a
consequence of it being a binary with respect to the level which is typical for single stars of the same mass
and evolutionary state.
397

Interacting Binary Systems include (IMA2 p. 672-3 and here
398
):
Algols Algol systems consist of two normal main sequence or subgiant stars in a semi-detached system.
They contribute mass to the ISM. Algol is an eclipsing binary consisting of a B8 and K2 star.
399

W Serpens Stars Active Algol systems.
RS Canum Venaticorum (RS CVn) Stars
400
Chromospherically active close binaries of F (cool) stars and
later spectral type components. They have large stellar spots which may cause luminosity to vary.
BY Draconis Stars Similar to RS Canum Venaticorum systems. The variability of BY Draconis is caused
by activity in the stellar photosphere called starspots, which are comparable to sunspots on the Sun, in
combination with rapid rotation that changes the viewing angle of the activity relative to the observer. This

397
http://articles.adsabs.harvard.edu/full/1995A%26A...298..482F
398
www.astro.caltech.edu/~george/ay20/Ay20-Lec11x.pdf
399
http://csep10.phys.utk.edu/astr162/lect/binaries/algol.html
400
http://en.wikipedia.org/wiki/RS_Canum_Venaticorum_variable
variation has an average periodicity of 3.8285 days, but the brightness also changes over the course of several
yearsdepending on the level of surface activity. Most observers believe that the primary star (A) is
responsible for the variability as the secondary produces only a third of the total luminosity from the system.
However, the spots may occur on both stars. Unlike the Sun, these spots may occur in the polar regions of
the stars
401

Cataclysmic Variables (CV) and Nova-like binaries
402
Have short orbital periods and contain a WD primary
component together with a cool M-type secondary that fills its Roche lobes. These irregularly and
dramatically increase in brightness as mass transfers lead to bursts of surface nuclear fusion.
W Ursae Majoris (W UMa) Contact Systems
403
Short orbital period (0.2 to 0.8 day) contact binaries
exhibiting very high levels of magnetic activity. Magnetic braking may cause these stars to coalesce.
X-ray Binaries
404
Binaries with a neutron star or, more rarely, a black hole primary. Accretion onto the
degenerate primary produces strong x-ray emissions. These provide evidence for the existence of black holes.
Zeta Aurigae ( Aur) Systems
405
Long orbital period binaries comprising interacting type G or K supergiant
and hot (~type B) companion. Although not originally interacting systems, they become so when the more
massive star evolves to become a supergiant.
VV Cephei Systems
406
Similar to Zeta Aurigae binaries with long orbital period, except a type M supergiant
replaces the G or K type supergiant.
Symbiotic Binaries Long orbital period (~200 to 1500 days) interacting binaries consisting of a cool type M
giant (sometimes a pulsating Mira-type variable) and a hot accreting companion such as a white dwarf, sub-
dwarf, or low-mass main sequence star. The hot star ionizes some of the material shed by the giant, and this
leads to the symbiotic's characteristic composite spectrum of nebular emission lines superimposed on
molecular absorption bands. Symbiotic binaries are valuable laboratories for the study of cool stellar winds,
ionized nebulae, wind-fed accretion processes and wind-wind collision dynamics.
407
Where the cool star fills
its Roche lobe, the system becomes a symbiotic Algol.
Barium and S-Star Binaries Long period binaries in which the originally more massive component evolved
into a white dwarf, transferring some of its nuclear-processed gas to a giant type K or M companion. (S
spectral type stars are intermediate between M and C stars.) Barium stars are spectral class G to K giants,
whose spectra indicate an overabundance of s-process elements by the presence of singly ionized barium,
Ba II, at 455.4nm. Barium stars also show enhanced spectral features of carbon, the bands of the
molecules CH, CN and C2.
408

Post-Common Envelope Binaries Usually consisting of a hot white dwarf or subdwarf, and a cooler
secondary, which have passed through the common envelope phase. The binary nuclei of planetary nebulae
are examples.

White Dwarfs in Semidetached Binaries
If the WD is the primary, these systems can exhibit dwarf novae, classical novae, or supernovae Type Ia.

401
http://en.wikipedia.org/wiki/BY_Draconis
402
Cataclysmic Variable Stars:
Szkody et al, Cataclysmic Variables From Sloan Digital Sky Survey. IV. The Fourth Year (2003), 2007,
http://iopscience.iop.org/1538-3881/129/5/2386/.../204549.text.html
http://en.wikipedia.org/wiki/Cataclysmic_variable_star
403
http://en.wikipedia.org/wiki/W_Ursae_Majoris
404
http://en.wikipedia.org/wiki/X-ray_binary
405
http://www.hposoft.com/EAur09/zeta%20Aurigae/zeta.html
406
http://en.wikipedia.org/wiki/VV_Cephei
407
http://www.astro.ex.ac.uk/people/th2/symbo.html
408
http://en.wikipedia.org/wiki/Barium_star
Cataclysmic Variables (CV)
This is a favorite research interest for Professor Szkody.
409
There are over 1600 known CV systems.
410
These
are subdivided into dwarf novae, classical novae, polars, and other less common types. CVs are binary star
systems that have a white dwarf and a normal star companion. They are typically smallthe entire binary
system is usually the size of the Earth-Moon systemwith an orbital period of 1 to 10 hours. The white dwarf
is often referred to as the primary star, and the normal star as the companion or the secondary star.
The companion star, a star that is normal, like our Sun, loses material onto the white dwarf via accretion.
Since the white dwarf is very dense, the gravitational potential energy is enormous, and some of it is
converted into X-rays during the accretion process. There are probably more than a million of these
cataclysmic variables in the [MW] galaxy, but only those near our Sun (several hundreds) have been studied
in X-rays so far. This is because CVs are fairly faint in X-rays; they are just above the coronal X-ray sources
and far below the X-ray binaries in terms of how powerful their X-ray emissions are.
411

Dwarf Novae
Dwarf novae have several subtypes, including Z Camelopardalis stars, SU Ursae Majoris stars, SS Cygni
stars . [Their] outbursts result from temporary increases in the rate of accretion onto the white dwarf,
caused by the additional material accreted onto the surface. This material must go through a violent
transition region called the boundary layer, which lies just above the surface of the white dwarf. Dwarf
novae outbursts are smaller in amplitude and higher in frequency than classical novae. The variable star U
Geminorum, or U Gem, is the prototype of dwarf novae. The brightness in the visible light of U Gem
increases by a hundredfold every 120 days or so, and returns to the original level after a week or two. Optical
astronomers have also recognized recurrent novae, which are eruptive behaviors that fall between the
definitions of classical and dwarf novae, and nova-like systems, which are stars that have similar spectra to
other types of CVs in the visual light, but have not been seen to erupt.
412
IMA2 p. 675 states that the
brightening takes place on the accretion disk, rather than at the WD surface, and that the outbursts are
caused by a sudden increase in the rate at which mass flows down through the accretion disk. The Tmax of
the accretion disk of Z Camelopardalis during an outburst is 4.4x10
4
K, the mass transfer rate is estimated at
1.3x10
-9
M yr
-1
(but lasting only days), and the total luminosity is 6.8 x 10
26
W, which is about twice L. The
durations of dwarf novae outbursts are typically 5 to 20 days, separated by 30 to 300 days of quiescence. The
mass transfer rate increases by a factor of about 10 to 100, similar to the brightness. (A 100x increase in
brightness is equivalent to a 5 magnitudes increase.) The mechanism of the irregular outbursts is unclear,
perhaps disk instabilities. (see IMA2 p. 676-8 for hypotheses) The graph below shows A portion of the SS
Cygni light curve from the AAVSO International Database displaying interchanging bouts of wide and narrow
outbursts.
413

409
http://www.astro.washington.edu/users/szkody
410
411
http://imagine.gsfc.nasa.gov/docs/science/know_l2/cataclysmic_variables.html
412
ibid.
413
http://www.aavso.org/vsots_sscyg
(Left) CV dwarf nova binary system Z Camelopardalis: Painting by Dale W. Bryner
414
shows
Roche lobe overflow and hot spot at region of impact.
(Right) Light curve segment of dwarf nova system SS Cygni, showing alternating wide and
narrow outbursts.
415

Polars and Miscellaneous CVs
Polars are also CVs. The AM Herculis stars are stars where the magnetic field of the white dwarf has locked
the binary system into nearly synchronous rotation. Matter streams straight down an accretion column fairly
constantly onto the white dwarf rather than forming a disc. They have fields of about 2000 T and emit X-ray
and UV light. The DQ Herculis types are also called intermediate polars. They have a weaker magnetic field
than AM Herculis stars, namely <1000 T. They form an accretion disk but do not exhibit synchronous
rotation. The visible light they emit is from nonrelativistic electrons spiraling along the magnetic field lines of
the accretion column. This light is called cyclotron (not synchrotron) radiation. It alternates between circular
and linear polarization, a phenomenon giving these stars the name polars. (IMA2 p 685 and here
416
)
Additional CV types include VY Sculptoris, AM Canum Venaticorum, and SW Sextantis.

Novae (Classical Nova)
The first nova witnessed was CK Vulpeculae in 1670. About 30 to 60 probably occur per year in the MW, but
only about 10 per year are identified due to obscuration by dust, etc. About 30 are seen annually in
Andromeda Galaxy and other nearby galaxies.
417
Classical novae have higher accretion rates than dwarf
novae. They have very large outbursts, of 7 to 20 magnitudes, caused by a runaway thermonuclear fusion of
material accreted onto the white dwarf. The amplitude of the outburst is the largest among CVs. Classical
nova outbursts are caused by sudden nuclear fusion of hydrogen-rich material on the surface of the white
dwarf...
418
Hydrogen fusion in a WD is possible only when fresh fusible fuel is accreted onto its surface.
In a nova, the primary WD averages 0.86 M, the secondary is usually a MS star of type G or later and less
massive than the WD. Orbital periods tend to be 78 to 720 minutes. During the outburst, the accretion disk
is more optically thick, suggesting more material is flowing through it. Doppler shifted emission lines appear
during the quiescent phase... (see IMA2 p. 673)
The peak temperature attained in the localized fusion process is as high as 10
8
K, and radiation pressure may
cause some of the accreted material to be blown away (the hydrodynamic ejection phase). A prolonged
equilibrium hydrostatic burning phase later sets in, which includes some CNO fusion. The hydrostatic
burning phase eventually ends and cooling begins, finally returning to the quiescent configuration, at which
point accretion resumes. However, it may take 10
4
to 10
5
years to build up a surface layer of sufficient size
(10
-4
M) to cause a recurrent thermonuclear fusion reaction and outburst. The mass transfer rate is
estimated at 9.0x10
-10
M yr
-1
(but lasts much longer than dwarf novae). (IMA2 p. 682, 684) Several
recurrent novae have been identified.

414
http://physics.weber.edu/carroll/
415
http://www.aavso.org/vsots_sscyg
416
417
http://en.wikipedia.org/wiki/Nova
418
http://imagine.gsfc.nasa.gov/docs/science/know_l2/cataclysmic_variables.html
The rise to peak luminosity for a nova is rapid, and takes only a few days. At peak brightness, these can have
luminosity of 10
5
L. During an outburst, V-band luminosity falls off faster than V+UV (bolometric)
luminosity. (IMA2 p. 682) The decline from the peak by 2 magnitudes takes a few weeks for a fast nova NA
and up to 100 days for a slow nova NB. Light output can be very irregular, however. (IMA2 p. 680 and
here
419
) Strope et al 2010 describe a variety of light curves for novae, designated as S, P, D, C, O, F, and J.
420

For example, the D class light curves stands for
Dust Dip: The defining feature of D class light
curves is a dust dip, which starts suddenly after
the initial decline. Dust dips appear as steep
drops in the brightness, although the depth of
this drop can vary. The dust dips then reach
some minimum brightness, followed by a
recovery in brightness to a fainter level than the
start of the dip, followed by a slow decline. The
classic D light curve is that of DQ Her [DQ
Herculis, a slow nova]. However, dust dips occur
with a very wide range of depths and durations.
The dust dip of V445 Pup is over ten magnitudes
in depth, taking the star to over four magnitudes
fainter than its normal quiescent level, with the
dip remaining nearly this deep at least until
2009... At the other extreme are shallow dips...
The seven light curve types are depicted in the
graphs to the right. The words for which the
letters S, P, D, C, O, F, and J are the first initial
are spelled out in the graph labels. Each nova
has its V-band light curve with an arbitrary
offset added so that the light curves are well
separated, with the shape being retained.
Curved line segments are superposed on the
binned and averaged magnitudes (diamonds) to
represent the fitted broken power law models.
The ejected gases pass through several stages.
The fireball expansion phase resembles an A or
F supergiant, and radiates at 600010000K.
(IMA2 p. 684) The initially optically dense
fireball (including the time of peak luminosity)
becomes less and less dense as it continues to
expand, eventually becoming transparent in the
optically thin phase. A limiting fireball photospheric radius of about 5x10
10
m is attained. (I have some
uncertainty about this statement.) After a few months, the carbon in the ejecta begins to coalesce as graphite
grains and dust forms, beginning the dust formation stage. The shell becomes optically thicker, and in the
visible light band, the central star may be hidden. However, bolometric luminosity remains high when the IR
emitted by the dust at about 900 K is included. The expanding shell may remain visible for many years after
the hydrostatic burning phase has ended. (IMA2 p. 682, 685)

419
420
Strope RJ, et al, Catalog of 93 Nova Light Curves: Classification and Properties,
http://arxiv.org/pdf/1004.3698v1

Evolution of an atypical nova-like red variable star V838 Monocerotis (V838 Mon)
and surrounding light echo (HST)
Regarding this sequence of images, this event was initially thought to be a nova. However, the Hubble Site
now states, The illumination of interstellar dust comes from the red supergiant star at the middle of the
image, which gave off a flashbulb-like pulse of light two years ago. V838 Mon is located about 20,000 light-
years away from Earth in the direction of the constellation Monoceros, placing the star at the outer edge of
our Milky Way galaxy... Astronomers do not fully understand the star's outburst. It was somewhat similar to
that of a nova, a more common stellar outburst... [But unlike a nova,] V838 Monocerotis did not expel its
outer layers. Instead, it grew enormously in size. Its surface temperature dropped to temperatures that were
not much hotter than a light bulb. This behavior of ballooning to an enormous size, but not losing its outer
layers, is very unusual and completely unlike an ordinary nova explosion... The outburst may represent a
transitory stage in a star's evolution that is rarely seen. The star has some similarities to highly unstable
aging stars called eruptive variables, which suddenly and unpredictably increase in brightness.
421
Although
not a typical nova, I could not pass up including these remarkably elegant HST images.

421
http://hubblesite.org/newscenter/archive/releases/2004/10/ image rearranged by MCM
Type Ia Supernovae
Unlike Type Ib, Ic, and II SNe, Type Ia SN vary relatively little in luminosity and rate of decline, etc., and
generally they vary in predictable ways that can be adjusted for. Aside from their intrinsic interest, these
serve an extremely important role as standard candles for the measurement of distances to the galaxies in
which they reside, and therefore of extragalactic and cosmological distances. They have provided critical
evidence that the rate of expansion of the universe has been accelerating ((t), the second derivative of the
cosmic scale factor, is >0) . This role, and the methods of analysis of light curves, luminosity, etc. are
161discussed in Chapter 27.
The typical peak luminosity expressed as absolute magnitude is MB MV 19.3 0.3 (IMA2 p. 686 and
erratum). In the scaled magnitude light curve (chapter 27), the time to peak is less than 20 days, and
averages about 19 days.
422

They lack H lines (like all Type I SNe), but exhibit strong Si II lines (including 615 nm), along with lines of
neutral and ionized O, Mg, S, Ca, and Fe. (IMA2 p. 687).
The mass ejected is reflected in P Cygni profiles (see Chapter 12) and the rates of expansion are shown by
blueshift to be as much as 10
7
m s
-1
(~0.1c).
As discussed also in Chapter 16 and 27, the model of the Type Ia SN starts with a WD in a binary systems that is approaching the
Chandrasekhar Limit for White Dwarfs of about 1.4 M. The companion overflows its Roche lobe and
contributes additional mass to the WD until the limit is reached or crossed, at which point a SN results.
A double degenerate model
423
depicts two WDs in a binary orbiting system. As they orbit, they give off
gravitational waves that remove energy and angular momentum, causing the stars to spiral in toward each
other. When sufficiently close, tidal forces break up the less massive larger WD in a few orbits and C-O rich
mass is dumped onto and merges with the other WD. Nuclear reactions begin deep in the interior and
eventually destroy the WD in a Type Ia SN (IMA2 p. 688)
An alternative and partially more favored model, however, is the single degenerate model. A WD has a
normal star as a companion, the latter overflows its Roche lobe, the WD attains the Chandrasekhar Limit,
and a SN Type Ia results, destroying at least the WD. This model would seem to lead to a more uniform mass
just at the Chandrasekhar Limit at the time of the SN. A helium flash may start the explosion, and eventually
the deeper C-O ignites in a deflagration (subsonic) or detonation (supersonic explosion), with internal
temperatures of billions of K.
In any event, the SN Type Ia ultimately results from the eruption of a C-O WD at or near the Chandrasekhar
limit of 1.4 M.
424
The energy released is 1210
44
J.
425
The mass ejected is reflected in P Cygni profiles (see
Chapter 12) and the rates of expansion are shown by blueshift to be as much as 10
7
m s
-1
. (IMA2 p. 689)
Another sources states that the star releases a shock wave in which matter is typically ejected at speeds on
the order of 5 to 20x10
6
m s
-1
, thus as high as 0.07c.
426

Neutron Stars and Black Holes in Binaries
After a core-collapse SN in one of the stars in a binary system (or an isolated NS capture), a NS or black hole
BH may be left in a binary system in which the compact degenerate star is orbiting the companion star.
When the companion overflows its Roche lobe, the gas infalling onto the compact star or its accretion disk is
associated with copious release of X-rays.

Supernova Explosion Impact on Binary Systems
After a core-collapse SN in a star M1 in a binary system, leaving a compact star with remnant mass MR, and
neglecting the direct blast effects on the companion star:

422
http://en.wikipedia.org/wiki/Supernova
423
See the Iben and Tutkov 1984 diagram earlier in this chapter
424
http://www.cfa.harvard.edu/news/2012/pr201214.html
425
http://en.wikipedia.org/wiki/Type_Ia_supernova
426
http://en.wikipedia.org/wiki/Type_Ia_supernova
If
/ (the ratio of post-SN M1mass to total initial binary mass 1/2),

the system is unbound
If
/ (the ratio of post-SN M1mass to total initial binary mass > 1/2),
the system becomes a NS or BH bound to the companion star M2.
IMA2 p. 690 states that for M2 >> M1, the latter case is likely (the description appears inconsistent however).

Capture of an Isolated Neutron Star
It is possible for a free NS resulting from a core-collapse SN to be subsequently captured by close approach to
another star. If the excess kinetic energy can be dissipated through tidal effects on the nondegenerate star, a
tidal capture results and the resulting close binary orbit will have an orbital period of hours to days. This is
most likely to happen in dense star regions, such as the center of globular clusters, and can be expected to
happen only 10 times in 10
10
years. Such a close binary X-ray system lasts only about 10
9
y, so X-ray
binaries created by this interaction are expected to be rare. Additional hypothesized capture mechanisms
include 3-body star interactions (in which the lightest star is ejected to carry away excess energy while the NS
is captured) and penetration of a NS into a giant star, the latter leaving a WD and NS binary with an orbital
period of about 10 minutes.

Binary X-ray Pulsars
These are also called X-ray pulsars. The first discovered, in 1962, was Scorpius X-1 (Sco X-1), which because
of its proximity (2800 pc) is the strongest X-ray source in the sky other than the Sun. Its X-ray luminosity is
about 2.310
31
W. It is a NS binary accreting matter at its Eddington accretion limit from a low-mass donor
companion star, and has an orbital period of around 18.9 hours.
427

Centaurus X-3 (Cen X-3) is another strong binary X-ray pulsar, the first binary pulsar to be discovered in X-
rays in 1971. The pulsar has a NS spin period of only 4.8 s and a binary orbital period of 2.09 d. This was
the first X-ray pulsar discovered. It consists of a 1.2 M NS that is regularly eclipsed by a massive 20.5 M
O-type supergiant companion which is feeding mass to the NS accretion disk.

Left graph: Typical X-ray emission profile of X-ray pulsar Cen X-3
for two full NS rotation periods of 4.8 s each
428

427
http://en.wikipedia.org/wiki/Scorpius_X-1
428
http://www.batse.msfc.nasa.gov/batse/pulsar/data/sources/cenx3.html image modified MCM
Right graph: Power spectrum showing variation in pulsar frequency
resulting from cyclical Doppler shifts as pulsar orbits its companion.
429

X-ray pulsars have X-ray luminosity up to the Eddington limit of 10
31
W and a Teff of 2x10
7
K, giving a peak
Wien emission = 0.15 nm, E = 8.27 keV, therefore in the X-ray band. (IMA2 p. 691).
X-ray pulsars also emit in radio frequencies, but this is not a prominent feature. Radio pulsars are mostly
single rapidly rotating NSs, whereas X-ray pulsars are binary systems, mostly accretion-powered and
consisting of a NS and an overflowing companion, and having slower and often more irregular rotation
periods.
430

X-ray pulsars like other NSs often have strong magnetic fields. These interact with the inflowing matter on
the accretion disk to divert mass away from the disk plane toward the poles. The disruption of radial inflow
and diversion takes place at a radius of ~0.5rA., where the Alfvn radius rA is given by:
)
/

and where BS = magnetic field at the surface, R is the NS surface radius. The channeled accretion, in a
typical strong magnetic surface field BS = 10
8
T, yields an Alfvn radius rA = 3.1x10
6
m and therefore a
disruption radius of about 1.5x10
6
m. This is 150x the 10
4
m NS radius R but still much less than the radius
at the outer disk margin (2rcirc) previously estimated. The accretion luminosity of the inner disk rises to the
Eddington limit, and the inner disk expands to up to 2R in height out of the disk plane. The process creates a
channel in which the particles follow the curving field lines and are funneled into the polar regions rather
than proceeding to the NS equator. In these polar regions, the infalling matter creates hot spots on the
circumpolar stellar surfaces, leading to a pulsating beam as the star rotates with the magnetic axis differing
from the rotational axis. (If the magnetic field is especially strong, the matter may never actually reach the NS
surface.) However, in contrast to a quasar, the X-rays are emitted over a large solid angle, especially in older
NSs where the magnetic field has become weakera sufficiently wide spread may suppress the periodic
variation entirely. See below about X-ray bursters.

Eclipsing Binary X-ray Systems
A NS with magnetic axis differing from rotational axis may be an eclipsing binary X-ray pulsar. A variety of
mechanisms for the eclipsing are proposed. IMA2 p. 695 states that there are about 20 known binary X-ray
pulsars. These rotate too fast to be WDs. The frequency of x-ray pulsar rotation increases with time (spin
period decreases), due to magnetic torque effects,
according to a complicated formula (IMA2 p. 695).
The NS of Hercules X-1 binary system rotates with spin
period 1.24 s and generates short-term pulses with this
period. Furthermore, it exhibits eclipses every 1.7 d
with the period of the binary orbit. It also undergoes an
almost periodic longer cycle consisting of 2 bright and 2
faint phases every 35 days. Excursions of 6 days in
both directions have been observed about these 35-d
cycles.
In the diagram to right, the short-term pulsations
resulting from NS rotation are shown as counts per bin
of 0.096 s over about 30 s.
431

429
http://chandra-ed.harvard.edu/clocks/clocks_A3.html image modified MCM
430
http://en.wikipedia.org/wiki/X-ray_pulsar
431
http://heasarc.gsfc.nasa.gov/docs/objects/binaries/herx1_spin.html
In the optical flux diagram to the right,
432
the 1.7 day
orbital period phase is depicted on the horizontal axis
and the longer 35 day cycle is on the vertical axis (X-ray
phase). Optical flux is a function of these two cycles
(plus of course the 1.24 s pulsar rotation.) The
backward-marching intensity ridges are evidence for
counter-precessing structure in the illumination of the
heated companion star [HZ Her]. In other words, it is
believed that the NS is precessing and that a tilted,
twisted accretion disk in retrograde (backward marching)
precession modulates the X-ray illumination and
reprocessed X-ray flux of companion HZ Her (and
therefore of the Earth), and this can explain the 35-day
intensity ridge cycle.
The diagram below
433
shows the gradual increase in
rotational frequency of the pulsar over an interval of
about 10,000 days. Analysis shows that this rate of
increase is much more compelling evidence for a NS as
the accreting star rather than a WD.

As the NS orbits, it changes distance from Earth in
a sinusoidal pattern and its relative velocity is also
varying sinusoidally. One or both effects (I am
unclear which) causes a cyclic variation in the measured
pulse period (arrival time) as a function of orbital phase.
In the graph
434
to the right, the time delay in s between
observed period and constant period is plotted against
orbital phase for the SMC X-1/Sanduleak 160 binary
system. This system has a 3.89 day orbital period, and
a pulse period = 0:715 s. It is said to be the most
luminous known X-ray pulsar.

432
Priedhorsky WC, Holt SS, Long-term cycles in cosmic X-ray sources, Space Science Reviews, vol. 45, no.
3-4, 1987, p. 291-348. http://dx.doi.org/10.1007/BF00171997
433
Bildstein L, et al, Observations of Accreting Pulsars, The Astrophysical Journal Supplement Series,
113:367408, 1997 December, http://dx.doi.org/10.1086/313060
434
Primini F, et al, Orbital elements and masses for the SMC X-1/Sanduleak 160 binary system,
Astrophysical Journal, vol. 210, Dec. 1, 1976, pt. 2, p. L71-L74. http://dx.doi.org/10.1086/182306
Magnetic X-ray Bursters (XRB)
If the magnetic field is sufficiently low, the H- and He-rich matter can with time accumulate sufficiently near
the surface to lead to release of bursts of thermonuclear energy lasting apparently only a few seconds, and
recurring in hours to days The fusion of H begins just below the surface, and He just below that. The He
burning is explosively rapid, releasing ~10
32
J in a few seconds, with transient surface T as high as 3x10
7
K.
An optical flash from the accretion disk may follow the X-ray burst by a few seconds. The rate of cooling
suggests blackbody radiation from an object of radius 10 km, thus a NS. Such X-ray pulsars are called X-ray
bursters XRBs. (IMA2 p. 691694, 696, and here
435
) Observationally, X-ray bursts are put into two distinct
categories, labeled Type I and Type II. A Type I X-ray burst has a sharp rise followed by a slow and gradual
decline of the luminosity profile. A Type II X-ray burst exhibits a quick pulse shape and may have many fast
bursts separated by minutes. However, only from two sources have Type II X-ray bursts been observed, and
so most X-ray bursts are assumed to be of Type I.
436
IMA2 p. 697 says there are about 50 known, but a list
of about 100 Type-I X-ray bursters can be seen here.
437
X-ray bursts are seen coming from Globular
Clusters, which are in the halo of our galaxy [the MW], and from sources along the plane of our galaxy
438

(whereas gamma ray bursts are distributed isotropically and are therefore mostly extragalactic).

Low-Mass X-ray Binaries (LMXB) and Massive X-ray Binaries (MXRB)
Mass determination in binary systems can be difficult, and in some cases only the ratio of the 2 stellar
masses is known (for this, inclination is irrelevant, Chapter 7) or a lower limit on the mass can be deduced
(uncertain due to unknown inclination). The X-ray Binaries are subdivided on the basis of the mass of the
donor companion M2 (not the compact star M1or total mass M1 + M2):
Low-Mass X-ray Binaries LMXBs: These are semi-detached binaries consisting of either a neutron star
or a black hole primary, and a low-mass secondary which is filling its critical Roche lobe. The companion
star is usually a late spectral type star with traditionally M2 1.5 M (sometimes stated as < 1 M or < 2
M), but can also be a WD or Spectral Type A star (Liu 2001) Typically, Loptical/LX-ray <0.1, and often
undetectable in visible light. They are powered by mass transfer via overflow at L1. They are seen in older
stellar populationsabout 25% are in globular clusters, where the probability of NS capture is higher.
Examples include Scorpius X-1 (Sco X-1), which has a 1.4 M NS accreting at its Eddington Limit,
439
and
A0620-00 (a BH candidate). CXOU J132518.2430304 is an example of a LMXB also with a suspected
stellar-sized BH.
440
They produce X-ray bursts, not regular pulses. The stars being low mass must be in
close proximity, so that orbital periods where known tend to be 11.4 min to 33.5 d (IMA2 p. 697). These
are rarethere were 150 known LMXBs in a 2001 article
441
and 100 are listed in a NASA May 2012
catalog.
442

Massive X-ray Binaries MXRBs: The companion star is higher mass, often a giant O or B star (that may
have survived a supernova), though often only a lower limit is known. Typically, Loptical/LX-ray > 1. The
traditionally defined MXRB mass is typically > 10.
443
About 1/2 are X-ray pulsars. The X-rays are

435
X-ray bursters XRBs:
Pter Mszros, The High Energy Universe: Ultra-High Energy Events in Astrophysics and Cosmology,
Cambridge University Press, 2010, p. 94-95
http://en.wikipedia.org/wiki/X-ray_burster
436
ibid.
437
http://www.sron.nl/~jeanz/bursterlist.html
438
http://imagine.gsfc.nasa.gov/docs/ask_astro/answers/961213b.html
439
http://en.wikipedia.org/wiki/Scorpius_X-1
440
Sivakoff GR, et al, A Transient Black Hole Low-Mass X-Ray Binary Candidate in Centaurus A, The
Astrophysical Journal, Volume 677, Issue 1, pp. L27-L30. http://dx.doi.org/10.1086/587960
441
Liu QZ, et al, A catalogue of low-mass X-ray binaries, A&A Volume 368, Number 3, March IV 2001,
http://dx.doi.org/10.1051/0004-6361:20010075
442
Low-Mass X-ray Binaries (LMXB) NASA Catalog:
http://heasarc.gsfc.nasa.gov/W3Browse/all/ritterlmxb.html and
http://www.mpa-garching.mpg.de/RKcat/
443
http://www-astro.physics.ox.ac.uk/~podsi/Flec_c1_4_c.pdf
harder, >15 keV. The stars are more widely spaced than in LMXBs, with orbital periods of 0.2 d to 580 d.
They are mostly found in the plane of the MW, where younger stars including O and B types can be found.
IMA2 p. 698 states that they have been almost entirely associated with NSs so far, but that a stellar mass
BH could also produce such X-ray bursts at the accretion disk. IMA2 concludes that Cygnus X-1, a bright
MXRB, is likely a BH candidate with a mass of about 14.8 M and a rotation rate at the event horizon of
>800 Hz.
444
Other articles such as this
445
regarding CXOU J132527.6-430023 report possible BHs as the
compact star in a MXRB, and these systems provide the best evidence for the existence of stellar-sized
black holes. To deduce a black hole, the mass of the compact star must be > 3 M, i.e., above the
TolmanOppenheimerVolkoff limit.
446

Intermediate-Mass X-ray Binaries IXRBs: some authors also use this term, but the exact bounds seems
to be fuzzy. Moreover, X-ray binaries in this range appear to be missing.

Fate of Binary Systems
As binary systems evolve, the low-mass secondary star in LMXBs typically evolves to a WD, with little
perturbation of the circular orbit. The higher mass secondary in a MXRB may explode as a SN, which
depending on relative masses as previously discussed, becomes a NS or BH. The result can be 2 NSs (or a
WD and a NS) in an elongated orbit, or the SN can kick the primary NS out of orbit to become a solitary free
NS. These free NSs, which are often pulsars, are mostly seen near the MW plane and often have high linear
velocities from the kick, from 400 to > 1000 m s
-1
.
Also, to recap earlier discussion in this chapter, if a WD in a binary has a close donor companion star, the
mass added to the WD may undergo fusion reactions leading to a nova, but this does not destroy the binary
system. In extreme cases, the WD can exceed the Chandrasekhar limit and trigger a supernova that destroys
the entire star.
447

Millisecond Radio Pulsars
A binary of 2 NSs, for which 1 is a pulsar, can be detected by cyclical variation in the measured periods of the
pulsar.
The first binary pulsar found, in 1974, was PSR 1913+16 (PSR B1913+16 or PSR J1915+1606). It has a spin
period of only 56 ms and is discussed below.
PSR 1937+214 , discovered in 1982, has a spin period of only 1.558 ms, or a rotation frequency of 642 Hz.
Despite the high rate of rotation, it is thought to be about 235 million years old. Although it is now solitary, it
was probably once the primary in a LMXB, in which accretion from the secondary sped up its rotation. Its
current rotation rate
448
is decreasing at a rate of about 110
-19
s s
-1
.
There appear to be 2 classes of binary pulsars. Those with high-mass companions evolve from MXRBs,
whereas those with low mass companions (such as WDs) evolve from LMXBs... (The details of this section,
IMA2 p. 702 are confusing and seem inconsistent).

Black Widow Pulsars
It turns out that there are few WDs in binary systems with a NS, and there may be a reason for this. An
example is PSR B1957+20, the so-called Black Widow Pulsar that is destroying its companion. This system

444
http://www.nasa.gov/mission_pages/chandra/multimedia/cygnusx1.html
445
Burke MJ, et al, A Transient Sub-Eddington Black Hole X-Ray Binary Candidate in the Dust Lanes of
Centaurus A, The Astrophysical Journal, Volume 749, Issue 2,
http://dx.doi.org/10.1088/0004-637X/749/2/112
446
http://en.wikipedia.org/wiki/Tolman%E2%80%93Oppenheimer%E2%80%93Volkoff_limit
447
http://en.wikipedia.org/wiki/Binary_star#Mass_transfer_and_accretion
448
http://en.wikipedia.org/wiki/PSR_B1937%2B21
is an eclipsing binary millisecond pulsar (period 1.7 ms) that orbits a brown dwarf
449
or WD (IMA2 p. 702)
companion. The orbital period is 9.2 hours and the eclipse duration is approximately 20 minutes. It is
believed that the companion star is being evaporated by ablation produced by photons and charged particles
from the neutron star: In black-widow systems, a millisecond pulsar is accompanied by a low-mass
companion [of a few hundredths of 1 M], which is bloated and strongly irradiated by the pulsar, leading to
outflows strong enough to eclipse the pulsar signal for significant fractions of the orbit. The irradiation causes
strong heating on the side of the companion facing the pulsar, and, as a result, strong orbital brightness
variations of the optical counterparts.
450
These authors estimate that the NS is 69 times heavier than the
companion, estimated at 2.4 M. The large mass also suggests that a large amount of mass was transferred
in the preceding phase as an X-ray binary, although this conclusion depends on the initial mass. However,
even if that were as high as [1.9 M], our measurements suggest that the pulsar has accreted about half a
solar mass. Within a few million years, this companion will have been destroyed (IMA2 p. 702).

Left image: This composite X-ray (red/white) and optical (green/blue) image reveals an
elongated cloud, or cocoon, of high-energy particles flowing behind the rapidly rotating pulsar,
B1957+20 (white point-like source). The pulsar, a.k.a. the "Black Widow" pulsar, is moving
through the galaxy at a speed of almost a million kilometers per hour. A bow shock wave due
to this motion is visible to optical telescopes, shown in this image as the greenish crescent
shape. The pressure behind the bow shock creates a second shock wave that sweeps the cloud
of high-energy particles back from the pulsar to form the cocoon... The Black Widow pulsar is
emitting intense high-energy radiation that appears to be destroying a companion star through
evaporation. It is one of a class of extremely rapid rotating neutron stars called millisecond
pulsars. The companion cannot be discerned. (NASA Chandra/ACIS = AXAF CCD Imaging
Spectrometer 2001)
451
Distance 5000 ly, width of image 1.2 arcmin.

Right Image: An artist's impression, viewed much closer, of the optical and X-ray emission of
the NS in the "Black Widow" pulsar B1957+20 and the ablation of the companion, which orbits
every 9.2 hrs. (NASA Chandra 2003)
452
The pulsar's orbit about the system barycentre has a
radius of 0.089 light seconds projected on to the line of sight.
453

449
http://en.wikipedia.org/wiki/Black_Widow_Pulsar
450
van Kerkwijk MH, et al, Evidence for a Massive Neutron Star from a Radial-Velocity Study of the
Companion to the Black-Widow Pulsar PSR B1957+20, The Astrophysical Journal Volume 728 Number 2,
http://dx.doi.org/10.1088/0004-637X/728/2/95
451
http://chandra.harvard.edu/photo/2003/b1957/
452
http://chandra.harvard.edu/photo/2003/b1957/b1957_ill2.jpg
453
http://www.nature.com/nature/journal/v333/n6170/abs/333237a0.html

Some of the ablated matter in a black widow system may condense to form one or more planets, a process
suggested to explain the 3 planets orbiting the NS PSR 1257+12.
454
Assuming the 6.2 ms pulsar has a
standard pulsar mass of 1.4 M, the derived masses of planets A, B and C are 0.03 M, 4.3 M, and 3.9M,
respectively.
455

The black widow PSR B1744-24A is another LMXB binary system in which it is thought that the secondary is
being ablated by the millisecond pulsar. It is located in globular cluster Terzan 5, the NS has rotation period
11.6 ms, and orbital period of 6536 s. The pulsar signal is at least partially eclipsed for about 1/2 of each
orbit, either with significant attenuation or complete absence of signal (depending on radio frequency). It is
presumed that the pulsar is enveloped or at least obscured by the large cloud of ablated material arising from
the companion. The eclipses are symmetrically centered at orbital phase 0.25, the stated point of inferior
conjunction (Nice and Thorsett 1992). Presumably the companion is interposed between NS and the Earth at
inferior conjunction. However, other sources state that binary conjunctions occur at orbital phase 0 or 0.5,
so there seems to be some inconsistency in defining binary orbital phase.) The eclipsing is total at 800 MHz
but only partial at 1667 MHz. This pulsar has a small but negative spin period derivative = 1.9 x 10
-20
s s
-1
,
meaning that it spinning up slightly (period is decreasing).
456
The findings are consistent with

Graphs of flux density of radio signals from PSR B1744-24A at 1667 MHz and 800 MHz
showing partial eclipsing at 1667 MHz (2 orbits shown) but total eclipsing at 800 MHz (4 orbits).
457

Double Neutron Star Binaries and Tests of General Relativity
Many double neutron star binaries have a pulsar as one of its NSs The other NS in double NS binaries is
usually not also a pulsar, and is detected only by its orbital effects.
Currently the only known exception is the double pulsar PSR J0737-3039 A/B,
458
which has two NS pulsars
of spin periods 23 ms and 2.8 s and masses of 1.4 and 1.25 M. The weaker pulsar is detectable when the
stronger is eclipsed.
459
(see also IMA2 p. 705, but the spin periods are a little confusing.)
The first Double Neutron Star Binary discovered was the Hulse-Taylor pulsar, PSR 1913+16 (aka PSR
B1913+16, PSR J1915+1606). It was discovered in 19744 by Russell Alan Hulse and Joseph Hooton Taylor,

454
455
Konacki M and Wolszczan A, Masses and Orbital Inclinations of Planets in the PSR B1257+12 System,
The Astrophysical Journal, 591:L147-L150, 2003 July 10, http://dx.doi.org/10.1088/0004-637X/728/2/95
456
Nice DJ and Thorsett SE, Pulsar PSR 1744-24A - Timing, eclipses, and the evolution of neutron star
binaries, Astrophysical Journal, Part 1, vol. 397, no. 1, p. 249-259, 1992,
http://articles.adsabs.harvard.edu/full/1992ApJ...397..249N Graph excerpted by MCM.
457
ibid.
458
http://en.wikipedia.org/wiki/PSR_J0737-3039
459
http://www.rigel.org.uk/work/pulsar.html
Jr.. Their analysis of the system, which strongly suggested that it was losing energy by giving off gravitational
waves in accordance with the general theory of relativity, earned them the 1993 Nobel Prize in Physics.
460
The
rotational period of the pulsar NS is only 56 ms, and the orbital period is 7.75 hr. The 2 NSs have a mass of
about 1.4 M each.
The rate of decrease of semimajor axis (orbital shrinkage) is 3.5 m y
-1
, and the rate of decrease of orbital
period is 0.0000765 sec per year or 2.4x10
-12
s s
-1
. Therefore, the orbital angular velocity of this binary
system is gradually speeding up as a result of shrinkage (inspiralling) of the orbit, precisely as predicted by
General Relativity with emission of gravitational quadrupole radiation. These results are therefore considered
to be a strong indirect confirmation of the existence of gravitational radiation. (IMA2 p. 704-5 and here
461
)
Here, the total power of gravitational radiation is estimated at 7.35 10
24
W, or about 0.02 L.
462
The total
orbital energy of the bound system correspondingly decreases (becomes more negative), even though the
orbital period decreases and the NSs orbit about each other more rapidly in angular velocity. The NSs will
eventually merge in about 300 million years.
463
The periastron separation in only 1.1 solar radii, and the rate
of periastron angular shift is 4.23 yr
-1
, much faster than Mercurys precession
464
of the perihelion. This
periastron angular shift rate is also predicted by General Relativity. (IMA2 p. 703-4)
Other double NS binary star systems include PSR 2127+11A M15 C in globular cluster M15),
465
B1534+12,
and J17562251.
466

Double neutron star pulsar binaries are superb natural laboratories in which to test the predictions of
General Relativity to a very high precision.
467
I have not included the formulas involved, but many are
included on IMA2 p. 7045 or the Wikipedia page on gravitational radiation. I hope someday to be able to
study the fascinating subject of General Relativity with more than superficial understanding.

Short-Hard Gamma Ray Bursts (GRB)
These are mentioned also in Chapter 15. Short-Hard GRBs are the result of compact object and are < 2 s in
duration, averaging 0.3 s. In other words, they are associated with NSNS mergers or NSblack hole
mergers but not with core-collapse SNe (the latter are associated with long-soft GRBs). In either case, a
single BH results (NASA 2005).
Perhaps the Hulse-Taylor double neutron star binary system will someday generate a Short-Hard GRB (IMA2
p. 706).

460
461
http://en.wikipedia.org/wiki/Gravitational_wave
462
463
Hulse-Taylor binary pulsar PSR 1913+16:
http://www.astro.cornell.edu/academics/courses/astro201/psr1913.htm
http://www.johnstonsarchive.net/relativity/binpulsar.html
464
Mercury Perihelion Precession:
The total observed precession of the Perihelion of Mercury is 574.6 arc-seconds per century relative to the
International Celestial Reference Frame ICRF, of which 43 arcsec is attributable to the effects of General
Relativity. The ICRF (see http://rorf.usno.navy.mil/ICRF/) is a quasi-inertial reference frame centered on the
barycenter of the solar system and defined by the measured radio positions of 212 extremely distant
extragalactic objects, mostly quasars, yielding a positional accuracy of about 1 mas in both directions.)
BepiColombo is a future mission planned to launch for Mercury in 2015. It is planned to test the general
relativity theory by measuring the parameters gamma and beta of the parameterized post-Newtonian
formalism with high accuracy.
http://sci.esa.int/science-e/www/area/index.cfm?fareaid=30 (ESA description of BepiColombo)
http://en.wikipedia.org/wiki/BepiColombo
http://en.wikipedia.org/wiki/Parameterized_post-Newtonian_formalism
http://prd.aps.org/abstract/PRD/v66/i8/e082001 (Testing general relativity with the BepiColombo radio
science experiment)
465
http://adsabs.harvard.edu/full/1996IAUS..165..279D
466
http://arxiv.org/abs/astro-ph/0508626
467
http://en.wikipedia.org/wiki/Tests_of_general_relativity
GRB 050709: A Short-Hard GRB was observed on July 9, 2005 (GRB 050709) which is helping to clarify the
nature of these transient events.
468
NASAs High Energy Transient Explorer (HETE-2) satellite was used to
make the initial discovery of the GRB. Follow-up observations of the HETE-discovered event were made using
ground-based telescopes, as well as the Chandra X-ray Observatory and Hubble Space Telescope. A team led
by MIT's George R. Ricker discovered a short GRB, designated GRB050709, lasting only 70 milliseconds on
July 9th [2009]. The GRB was detected by NASA's High-Energy Transient Explorer (HETE-2), the first satellite
dedicated to GRBs. This particular short burst provides a long-sought nexus, enabling detection of the
prompt emission and its afterglow, from the gamma-ray band to the optical, for the very first time... HETEs
accurate localization of the burst allowed NASAs Chandra X-ray Observatory, Hubble Space Telescope and
ground-based telescopes to identify the bursts X-ray afterglow, and, for the first time, its optical afterglow,
which provided the clues needed to track the burst to its host galaxy. The distinctive signature is that of two
neutron stars or a neutron star and a black hole merging, followed by a colossal explosion [rather than a
supernova].
469

Left Graphs: The short X-ray and gamma-ray spike associated with GRB 050709 (July 9, 2005),
shown in multiple X-ray and gamma ray energy ranges over about 470 s. An X-ray afterglow
bump is also apparent in 2-10 keV. Time history observed by WXM [HETE-2 Wide Field X-ray
Monitor] in the 210 keV energy band (a) and in the 225 keV energy band (b); time history
observed by FREGATE [HETE-2 French Gamma Telescope] in the 640 keV energy band (c) and
in the 30400 keV energy band (d). The event is a short-hard [X-ray] spike of duration T90 =
22050 ms in the 225 keV energy band and 7010 ms in the 30400 keV energy band,
followed, ~25 seconds later by a long-soft bump of duration T90 = 130 7 s in the 225 keV
energy band...

Right Graphs: Same, shorter time scale. Time histories observed by the WXM in the 210 keV

468
http://www.nasa.gov/mission_pages/swift/bursts/short_burst_oct5.html
469
http://web.mit.edu/newsoffice/2005/gamma-ray.html
energy band (a) and the 1025 keV band (b); and by FREGATE in the 630 keV energy band (c),
the 3085 keV energy band (d), and the 85400 keV energy band (e), plotted in 5 ms time bins.
The pulse has a duration of T90=22050 ms in the 225 keV energy band and 7010 ms in the
30400 keV energy band, and exhibits no detectable emission before T = 0 or after T=400 ms,
confirming the short, hard nature of the pulse. (Villasenor 2005)
470

The location of the X-ray and optical afterglow was initially found to lie in a circle of 90% probability having
radius of 1.34 arcminutes established by the Soft X-Ray Camera (SXC) for the GRB, but the localization was
subsequently improved, as noted below.

Left X-ray Image (detail): HST and Chandra X-ray Observatory images of the afterglow and
environs of GRB 050709. ... The Chandra (0.38.0 keV) image of the field from our observation
of 2005 12.5 July UT [3 days after the July 9 GRB]. The large red circle is the HETE
localization region, 81 arcsec in radius... A red ellipse indicates [an unrelated object]... The
bright point source in the boxed region [box is about 1 arcmin on a side] is the afterglow of
GRB 050709... Our Chandra [X-ray] afterglow candidate was found to be coincident with a
point-like optical source.... Apparently the bright white spot is X-ray afterglow, not optical.

Right Image: Close-up of the region surrounding the X-ray afterglow, in a co-addition of all our
HST [thus optical] data; the red circle is the Chandra localization region, 0.5 arcsec in radius.
A point source is visible within this region; the source is observed to fade over the course of our
HST observations [see images below], and we identify it as the optical afterglow of GRB
050709. The irregular galaxy to its west [i.e., to the right] is the proposed z =0.16 host galaxy...

470
Villasenor JS et al, Discovery of the short g-ray burst GRB 050709, Nature Vol 437, 6 October 2005,
doi:10.1038/nature04213
We find [in] a best-fit solution ... the afterglows 1.38-arcsec offset from the brightest region of
this galaxy corresponds to 3.8 kpc in projection.
471

Image above: Observations of the GRB 050709 afterglow and illustrative models (curves).
X-ray = red, Optical = green, Radio = blue.
...These multi-wavelength observations can be marginally accommodated within a standard
external-shock afterglow model... (see Fox 2005 for details) Note the log-log scaling.

Optical afterglow imaging: These Hubble Space Telescope images show the fading afterglow and host galaxy
of the HETE short burst of July 9, 2005. The images are taken 5.6, 9.8, 18.6, and 34.7 days after the burst,
respectively. The bright, point-like afterglow is located to the left, and fades away over the course of the
month following the burst. The colors indicate the intensity of red light (814 nm) as seen by the Advanced
Camera for Surveys instrument on HST.
472
The host galaxy is described as a bright elliptical galaxy.
473

471
Fox DB, et al, The afterglow of GRB 050709 and the nature of the short-hard g-ray bursts, Nature Vol
437-6 October 2005, doi:10.1038/nature04189
472
Images cropped and assembled by MCM, from
http://www.nasa.gov/mission_pages/swift/bursts/short_burst_oct5.html
473
ibid.
The Solar System & Planetary Systems (Chapters 19 through 23, omitted)

These chapters have not yet been read in my astronomy courses and are not summarized here.

The Milky Way Galaxy (Chapter 24)

History of Discoveries
The Milky Way Galaxy (MW) is named after its milky appearance. In fact, the word galaxy derives from
Hellenistic Greek (short for =milky circle) which came from ancient Greek ...
=milk (OED). It is our galaxy, and therefore often called simply the galaxy. Galileo in 1610 first observed
that it is made up of myriad stars. Immanuel Kant (17241804) in 1755 suggested it is a large stellar disk,
with our Sun only a component of it, and by extrapolation that other elliptical nebulae might be additional
extremely distant island universes.
William Herschel (17381822) produced the first attempts to map the MW 1785. The model of Jacobus
Kapteyn (18511922) assumed the galaxy is a flattened spheroidal system and estimated stellar number
density. Harlow Shapley (18851972) came up with a much larger (in fact excessively large) estimate of
galaxy size, about 100 pc diameter, based on analysis of globular clusters, and he placed the Sun farther from
the center. Early estimates of sizes were in error due to presence of unrecognized or hard to quantify
interstellar extinction. Its effects were apparent for instance from the paucity of globular clusters within 10
of the galactic plane.

Interstellar Extinction
This topic has been previously introduced. In the Milky Way disk, the typical extinction rate of visible light is
about 1 magnitude kpc
-1
, though obviously worse in dust lanes. The degree of extinction can be estimated by
the amount of interstellar reddening. Extinction is much less for infrared.

Differential and Integrated Star Counts
Modern astronomy places great emphasis on large datasets of stars (such as SDSS,
474
which by 2012 is said
to have made observations on 500 million objects). The datasets are categorized by spectral types, size,
luminosity, etc. The data are superimposed on models of star distributions, extinction, variations in
composition, etc. that are iteratively fine-tuned until a best fit is found.
A Differential Star Count AM expresses the number of stars visible in selected regions of the sky counter over
a specified apparent magnitude scale. More precisely, IMA2 p 880 gives a defining formula for AM, the
number of stars having attribute(s) S and with an absolute magnitude between M and M + dM that are found
within a particular solid angle and that have apparent magnitude between m and m + dm.
An Integrated Star Count NM expresses the number of stars having attribute(s) S and with an absolute
magnitude between M and M + dM that are found within a particular cone defined by a solid angle and
extending from the observer to a distance d. The distance range criterion is commonly replaced by an
apparent magnitude range to give an alternate form of the Integrated Star Count,
, where the bar

designates a function of m rather than of d.

474
http://en.wikipedia.org/wiki/Sloan_Digital_Sky_Survey
Recent and Future Astrometric Surveys of the Milky Way and Other Galaxies
According to ZXI lecture 2 and an article he co-authored
475
, use of the all-important SDSS dataset in a
photometric survey employing color-color diagrams (for example, g-r plotted against u-g), yields valuable
automated data conveying much information. In the following image from that article, a locus of binary stars
not previously identified was demonstrated:

475
Smoli V, Ivezi Z [ZXI], et al, A Second Stellar Color Locus: a Bridge from White Dwarfs to M stars , The
Astrophysical Journal Letters Volume 615 Number 2, 2004.
In a 2008 review article ZXI also co-authored, the SDSS dataset of the Milky Way was used to establish
photometric distances to map 3-D number densities of stars in tomography-like cross sections. The authors
show the general structure of the MW, with a thin disk transitioning to thick disk, an oblate halo, and a
number of localized overdensities including a Virgo overdensity in the halo toward the constellation of Virgo.
(The locations studied only covered regions more than 3 4 kpc from the Galactic center, where the bulge
contribution is negligible.)
476

The following images also help to demonstrate the value and future potential of computer-intensive number
crunching of dataset like the SDSS for galaxies. Although I have not explored or understood the details, ZXI
states (Lecture 3) that in the upper plots (of concentration index vs. u-r), the galaxies without emission lines
(left image) are concentrated in the upper right red region and represent ellipticals, whereas the galaxies with
emission lines (right image) are concentrated in the lower left blue region and represent spirals and AGNs. In
the lower image, a BPT diagram (Baldwin, Phillips & Terlevich 1981)
477
plots luminosity ratios including two
different forbidden transitions ([O III] and [N II]) for emission line galaxies. The article, which ZXI
co-authored, states Emission-line galaxies can be separated into three groups according to their position in
the BPT diagram: AGNs, star-forming and unknown, using the separation boundaries outlined by the dashed
lines.
478
This technique clearly has value to help in identifying AGNs in an automated manner.

476
Juri M, Ivezi [ZXI], et al, The Milky Way Tomography with SDSS. I. Stellar Number Density
Distribution , The Astrophysical Journal Volume 673 Number 2, 2008
477
Baldwin JA; Phillips MM; Terlevich R, Classification parameters for the emission-line spectra of
extragalactic objects, Astronomical Society of the Pacific, Publications, vol. 93, Feb.-Mar. 1981, p. 5-19.
478
Obri M, Ivezi [ZXI], et al, Panchromatic properties of 99 000 galaxies detected by SDSS, and (some by)
ROSAT, GALEX, 2MASS, IRAS, GB6, FIRST, NVSS and WENSS surveys, Mon. Not. R. Astron. Soc. 370, 1677
1698 (2006)
Professor Ivezi is designated System Scientist for the upcoming Large Synoptic Survey Telescope (LSST)
project,
479
which should provide further riches from surveying several billions stars. It will be located at Cerro
Pachn in Chile and is scheduled to see first light in 2019. Its website states, it will produce a 6-band (0.3-
1.1 micron) wide-field deep astronomical survey of over 20,000 square degrees of the southern sky using an
8.4-meter ground-based telescope. Each patch of sky will be visited about 1000 times in ten years. The LSST
leverages innovative technology in all subsystems: the camera (3200 Megapixels)..., telescope ..., and data
management (30 terabytes of data nightly, nearly instant alerts issued for objects that change in position or
brightness)...
The space-based GAIA Mission of the ESA is also highly ambitious and eagerly anticipated. It will be placed
in orbit at the L2 Lagrangian point, which lies at a distance of 1.5 million kilometres from the Earth in the
anti-Sun direction and co-rotates with the Earth in its 1year orbit around the Sun.
480

Perhaps I will have more time eventually to delve deeper into these new and impressive areas of research.

Milky Way Fundamentals, Morphology, and Coordinates

Our galaxy is hard to study because we are embedded in it, there is substantial extinction close to the galactic
plane, and because we are on a moving platform moving in a complex way relative to the galactic center.
The Milky Way is a barred spiral galaxy, type SBc (or SBbcIII). It has approximately 300 x 10
9
stars, a total
mass of 1.0 to 1.5 x 10
12
M, and a bolometric luminosity of 2 to 3 x 10
10
L (based on an estimate of total
luminosity LMW = 10
44
ergs/s). The oldest known star is 13.2 Gy.
481
The average MW stellar mass is about 0.7
M, thus dwarf stars smaller than the Sun predominate (IMA2 p. 887).
As shown in the diagram to follow, the full galactic disk (stars, dust, and gas) is about 50 kpc = 163,000 ly in
diameter, although the disk is thought to be somewhat elliptical (0.9 minor/major) with the Sun near a major
axis.

479
http://www.lsst.org/lsst/
480
481
Anna Frebel et al, Discovery of HE 15230901, a Strongly r-Process-enhanced Metal-poor Star with
Detected Uranium, The Astrophysical Journal Letters Volume 660 Number 2, 2007 ApJ 660

Gross Morphology of the Milky Way(
482
)

Galactic Coordinate System and Cylindrical Coordinate System
The central plane of the Milky Way is oriented at 62.87 with respect to the plane of the celestial equator. In a
particular view, the north celestial pole NCP, north ecliptic pole NEP, and north galactic pole NGP are all
in the same region of the celestial sphere. Looking down from the NGP, the Milky Way is seen to rotate
clockwise with respect to the celestial sphere (the opposite of the Earths rotation about the Sun).
Galactic coordinate angles are measured with the Sun at the vertex of the angle (even though the Sun is not
at the galactic center). The axis pointing from the Sun to the center of the galaxy, ~8 kpc away, defines 0
galactic longitude (=0). The galactic longitude for an object is measured in an easterly direction
(counterclockwise looking down from NGP) from the direction of the 0 galactic longitude to the intersection
with the galactic equator of a plane oriented perpendicular to the galactic plane and passing through the Sun
and the object (but not usually through the NGP!). The galactic latitude b for an object is the angle in degrees,
positive for north and negative for south, between the object and the galactic plane with Sun at the vertex and
lying again in a plane holding the Sun and the object and perpendicular to the galactic plane.

482
from http://faculty.physics.tamu.edu/papovich/courses/milkyway2.pdf , which derives from image in
IMA2 p. 883

Comparison of Galactic, Ecliptic, and Celestial Coordinate Systems:
NCP=north celestial pole (blue), NEP=north ecliptic pole (gold), NGP=north galactic pole (green)
Celestial equator (blue), Ecliptic (gold), galactic equator (black)
Rotation of galaxy shown wrt fixed celestial sphere.
(Rotation annotated by MCM)
483

483
MCM modified from http://en.wikipedia.org/wiki/Celestial_coordinate_system

Diagram showing how latitude and longitude are measured with the Sun at the vertex.
484

Galactic motions (kinematics) are measured in a cylindrical coordinate system, centered on the galactic center
and using coordinates R, , and z. Here, is the azimuthal-type angle pointing in the direction of rotation
and with the Sun at 0, R is the radius from the galactic center to the objects projection on the galactic plane,
and z is vertical distance from the galactic plane.

Thick and Thin Disk
The disk is subdivided into the thin disk and the thick disk, the latter being much lower in M, L, and star
number density. Disk heights and lengths are expressed by scale heights and scale lengths. Scale height
is defined as the vertical z distance from the central Galactic midplane over which a quantity such as the
number density decreases from the typically higher central value by a factor of 1/e = 0.37 (i.e., the final value
is 1/e times the initial value). At the central axis of the disk, the thin disk has a scale height of 350 pc =
1140 ly, whereas the thick disk is ~1000 pc = 3300 ly. The disk scale length in the radial direction is about
2.25 kpc. The central star number density is n0 0.02 stars pc
-3
. IMA2 p. 883 gives a formula for star
number density n(z, R) in the disk, based on these scale heights and scale length.
The thin disk consists of mostly younger stars, dust and gas. The thicker disk consists of older stars and has
a stellar number density of about 9% of that in the Galactic midplane. Comparative values for M, L, R,
[Fe/H], Age, etc. are given on IMA2 p. 884.
Disk stars rotate in near-circular orbits about the galactic center and stay confined to the disks.

Spiral Bands
The spiral structure so readily apparent in other galaxies (such as Andromeda M31) is more difficult to
discern in the MW because we are embedded in it. The arms are the site of new star formation and exhibit
more blue light as a result of greater numbers of these young, hot, bright massive stars (though not
necessarily much greater number density of stars overall). Spiral arms are less conspicuous in red light. The
Sun is close to but not in the Orion-Cygnus arm, aka the Orion arm or Orion spur.

484
www.astro.sunysb.edu/metchev/AST443/lecture2.pdf (Stanimir Metchev)

A Hypothetical Map of the Milky Way showing Spiral Bands.
Note also that the Orion Arm is not a major spiral arm, but only an enhancement of stars and gas between
the Sagittarius and Perseus arms. It should be emphasized that there are almost as many stars between the
spiral arms as in the spiral arms. The reason why the arms of spiral galaxies are so prominent is that the
brightest stars are found in the spiral arms. Spiral arms are the major regions of star formation in spiral
galaxies and this is where most of the major nebulae are found.
485

Interstellar Gas, Dust, and Peripheral Warp
The 21-cm microwave and IR emissions of CO are useful for assessing H I and H2 in the central plane and the
distant reaches of the MW. (IMA2 p. 889) At R > 12 kpc, the scale height of H I increases and a pronounced
warp in the z-axis-distribution of H I becomes apparent by R = 13.6 kpc.
486

485
http://www.atlasoftheuniverse.com/milkyway.html
486
Burton WB, te Lintel Hekkert P, The shape of the outer-Galaxy H I layer. I - Atlas of volume densities in
cuts through a composite, galactocentric data cube, Astronomy and Astrophysics Supplement Series vol. 65,
no. 3, Sept. 1986, p. 427-463
There is evidence for high-velocity clouds of hydrogen, and a Galactic Fountain model has been suggested,
perhaps arising as SN ejecta. There is a very hot tenuous gas at distances up to 70 kpc from the galactic
center, apparent in other galaxies from O VI absorption lines. This hot gas is estimated to have T > 10
6
K. By
analogy to the Sun, it is called a coronal gas.

Disruption of Satellite Galaxies
The Magellanic Stream is a narrow but long band of H I radio emission stretching more than 180 degrees
across the Southern Hemisphere and trailing the Magellanic Clouds. It is proposed that this stream is a
result of tidal encounters of the Large and Small Magellanic Clouds with the MW some 200 Ma. (IMA2 p. 891)

Magellanic Stream Radio Emission Superimposed on All-sky Visible Light Image.
Wide angle radio imagesincluding those from the Byrd Green Bank Telescopehave shown
that the Magellanic Stream is longer and older than previously thought, perhaps as old as 2.5
billion years. These observations bolster a third possible origin for the streamthat the Large
and Small Magellanic Cloud galaxies [SMC and LMC] once passed so close to each other that
gravitational tides triggered a burst of star formation that left the stream. Pictured above
digitally superposed on a recently-completed all-sky image in visible light, the radio emission of
[the] Magellanic stream is shown in false color pink extending across the sky and ending at the
two Magellanic galaxies on the lower right. (NASA)
487

Other possible former or current satellite galaxies that may have interacted with the MW include the
Sagittarius dwarf spheroidal galaxy (leading to the Sagittarius Stream), a group of globular clusters in Canis
Major, and the unusual globular cluster Centauri.

Galactic Bulge
This feature, which is better seen in IR, is thought to be an independent component of the MW (and of many
spiral galaxies). It has a mass of about 10
10
M, and B-band luminosity of 3x10
9
L. and M/L 3 M/L.
The somewhat boxy bulge is apparent in COBE IR views (IMA2 p. 892).
As with other galactic features, the ill-defined boundaries of the bulge may be delineated by a threshold scale
height (distance for quantity to drop to 1/e = 0.37 of the highest value), and for the MW this is 100 to 500 pc
depending on star type.

487
http://apod.nasa.gov/apod/ap100125.html
The quantity assessed to define its boundary is typically surface brightness I. (See earlier discussion of
surface brightness) An r
1/4
relationship for I has been empirically found (IMA2 p. 892), using the effective
radius re and Ie as this radius as normalization values:
o
]
[
(
]
,

here expressed in units of L per pc
-2
. This is termed de Vaucouleurs' law
488
, a special case of Sersics
(Srsics) law with Sersic index n=4 (discussed later). By IRAS, the effective radius re is found to be 0.7 kpc.
The chemical characteristics and metallicity of stars in the bulge are discussed at IMA2 p. 893.
The stars have largely random orbits but with some net rotation about the Galactic center. There is a ring of
gas and dust near the center. (ZXI lecture 2).

Central Bar and 3-kpc Expanding Arm
This has been found in the bulge. It has a half-length (radius) of ~4.4 kpc, is somewhat thicker in the central
plane of the galaxy, and is oriented at 44 with respect to the Sun to Galaxy center line. An inner expanding
arm at 3 kpc seen near and far appears to be related to the bar.
489

Stellar Galactic Halo
This roughly spherical feature of visible (luminous or demonstrable) matter extends about 50 kpc from the
disk, thus about the same as the diameter of the disk. It consists of globular clusters of stars and field stars
(stars that are not part of a cluster and that often have a high velocity). Two populations of globular clusters
are apparent: older metal-poor globular clusters ([Fe/H] < -0.8), which extend in a nearly spherical
distribution, and younger clusters with [Fe/H] > -0.8 which occupy a flatter distribution (see IMA2 p. 894 ff
for details). Some of the distant clusters may have been captured as clusters of dwarf galaxies. It contains
little gas and dust, and no recent star formation. Unlike the disk stars (which orbit around the galactic center
in the disk plane), the halo stars have randomly distributed elliptical orbits in three dimensions.
Although it was previously thought that there was no obvious substructure in the halo, structural features
such as the Sagittarius Stream, Virgo Overdensity, Monoceros Ring, and Orphan Stream have been identified
with SDSS data (ZXI lecture 2 and here
490
).
The halo is of interest because it exhibits merger history and because the gravitational potential becomes
dominated by the dark matter halo (ZXI Lecture 2b).

Total Galactic Luminosity and Mass (not including dark matter)
The estimated total MW luminosity in the B-band (blue band) is LB, MW tot is ~2.3 10
10
L and total bolometric
luminosity, including IR, is Lbol, MW tot is ~3.6 10
10
L. (As usual, I believe that the total bolometric value of L
is used for the comparison, not the blue band solar luminosity, Lblue, .) The total mass MMW is ~9 10
10
M.

Dark Matter Halo
This is inferred by the unexpectedly flat distribution of star and gas rotational velocities beyond R0 (8 kpc)
from the center, and may also contribute to the peripheral warp mentioned. It extends to 230 kpc or farther,
with an empirical density distribution of

488
http://en.wikipedia.org/wiki/De_Vaucouleurs%27_law
489
http://www.cfa.harvard.edu/mmw/Far_3kpc.html
490
http://www.sdss.org/news/releases/20060508.mergers.html

/ +/

where (r) is density at 3-D radius r, and 0 and a are fitted values (IMA2 p. 896, see later discussion). The
dark matter halo comprises about 95% of the mass of the MW. The nature of this dark matter is actively
debated and under investigation (and not summarized here). Possibilities include nonbaryonic weakly
interacting massive particles (WIMPS), neutrinos, neutralinos, and massive compact halo objects (MACHOS,
such as black holes, neutron stars, and white or brown dwarfs).

Galactic Magnetic Field
Typical MW magnetic fields (based on polarization by aligned interstellar grains and by Zeeman splitting) are
0.4 nT in the spiral arms, 1000 nT in the galactic center, and perhaps 0.04 nT in the stellar halo. By
comparison, the Earth magnetic field is about 5x10
-5
T = 50 microT = 50,000 nT = 0.5 gauss.
Despite the low field strength, the energy density from the magnetic field is comparable to that of the thermal
energy, at least in the disk.

Age-Metallicity Relation
The Population I stars are relatively metal-rich with Z~0.02, whereas Population II are metal-poor with
Z~0.001. Metallicity is also expressed by [Fe/H], and [O/H], as previously discussed, because Fe lines are
readily observed in spectra. During a supernova explosion, particularly Type Ia, Fe is ejected enriching the
ISM and metallicity of future stars that form. Iron content then correlates with stellar age, greater metallicity
being found in younger stars (such as the Sun for which [Fe/H]=0.0 by definition), and lower metallicity is
seen in old extremely metal-poor stars (with values down to 5.4). Metallicity has been used to show that thin
disk stars forming ~8 Gya and more recently are younger by 2 or 3 billion years than thick disk stars, which
formed ~10 - 11 Gya. (IMA2 p. 886)
This is the age-metallicity relation, but it is not always reliable. Type IA SN do not occur until about 10
9
years
after star formation, and mixing of SN ejecta in the ISM is not uniform. Core-collapse SN produce more
[O/H], which may be helpful. Age estimation is further complicated by the strong dependence on distance
modulus.

Kinematics of the Milky Way

(Highly abbreviated and selected)
As noted, galactic kinematic motions are measured in a cylindrical coordinate system, centered on the
galactic center and using coordinates R, , and z. Here, is the azimuthal-type angle pointing in the
direction of rotation and with the Sun at 0, R is the radius from the galactic center to the objects projection
on the galactic plane, and z is vertical distance from the galactic plane. (The symbol r is customarily reserved
for 3-D distance to the object from the center.) Object velocity components for R, , and z are expressed as ,
, and Z, respectively.
Formulas for conversion to/from the galactic coordinate system are given on IMA2 p. 901 ff, including a graph
for making these conversions.
Observations measured from the moving Earth must be corrected for peculiar motions and the Local
Standards of Rest LSR. The LSR has an orbital speed of 0(R0) = 220 km s
-1
. The velocity of a star relative to
the LSR is known as the stars peculiar velocity. The peculiar velocity is given by the V = (u,v,w), with
components representing components VR,V and VZ. These and other complex kinematic matters are
discussed in IMA2 p. 901913. The Suns own peculiar velocity relative to the LSR is the solar motion.
For a star, plotting one component of peculiar velocity against another, one can plot velocity ellipsoids. These
can be used to show that the oldest metal-poor stars in the Galaxy have the widest range of peculiar velocities
in all three coordinates, whereas younger metal-rich stars have smaller peculiar velocities and are found in
the thin disk. Stars with large |w| are passing through the solar neighborhood on trajectories that will carry
them to great distances above and below the disk.
The amount of variation in each component of velocity for a population of stars is expressed as the standard
deviation of the velocity distribution, given by

]
/

When <u> is zero, the quantity here computed is

/

and is termed the velocity dispersion.
The Milky Way is moving in the direction of = 10.5, = 24 (J2000, near the center of Hydra)
at 552 km/s with respect to the CMB rest frame.
491

The Sun is moving toward the solar apex, opposite from the solar antapex. Objects in eccentric orbits about
the center are said to attain the apogalacticon (greatest distance from the center) and the perigalacticon
(closest distance).
It takes about 250 million years for the Sun to make a full orbit around the Milky Way.

Differential Galaxy Rotation and Oorts Contributions
The galaxy is rotating with differential rotation, as first analyzed by Jan Oort (19001992). The angular
velocity of an object is given by (R) = (R)/R where (R) is linear orbital velocity. Oort derived formulas
incorporating Oort Constants A and B which give vr Ad sin 2l and vt Ad cos 2l + Bd. (Here, l is galactic
longitude, d is distance from Sun to star, see IMA2 p. 911).

491
http://en.wikipedia.org/wiki/Milky_Way

Diagram of the various rotation curves in a galaxy
492

The Oort constants help one understand the types of possible galactic rotation. Solid Body rotation is self-
evident, the angular velocity (R) is constant with R and rotational velocity (R) rises steadily with increasing
R. There is no shear motion, A = 0, and the vorticity is just the angular rotation B = -. This gives the upper
limit or fastest an object can rotate for a given radius. This model fits the observed data only in the inner
regions of the galaxy (see graph).
In the case of Keplerian rotation, where v = (GM/r)
1/2
, observed rotation values of Oort constants do not
match the derived values, so the Keplerian model appears invalid. It can however be thought of as providing
the minimum velocity an object can have in a stable orbit.
The flat rotation curve in which linear rotational velocity (R) is constant and independent of radius yields
Oort constants that fit remarkably close to observed data in the solar neighborhood and father out. It
provides a value intermediate between rigid and Keplerian motion.

Hydrogen 21-cm Line as a Probe of Galactic Structure
Because of light extinction, this radio or microwave line from H I is very useful in exploring the entire galactic
disk. However, it is difficult to determine distance d with it. The technique is discussed on IMA2 p. 914.

Estimating Distances in the Milky Way
The distances to MW stars for galaxy analysis are most reliably estimated with Cepheid Variables.
The moving cluster method is also important. Early work was done especially with the Hyades open cluster
of stars, a cluster in the MW at about 46 - 47 pc.
493
The directions of stellar motions over time may be
deduced, after removing the effects of the Suns peculiar motion. When plotted in equatorial coordinates, the
star motions appear to point to a convergent point. The distance to the cluster may be determined from

492
http://en.wikipedia.org/wiki/Oort_constants
493
http://en.wikipedia.org/wiki/Hyades_%28star_cluster%29
knowledge of the direction to the convergent point, the average radial velocity of the cluster members, and the
average of their proper motions (IMA2 p. 921).
Once a distance to a cluster is deduced, one may calibrate the main sequence. By comparing the apparent
magnitudes of other cluster H-R diagram main sequences, it is possible to find the distance modulus of these
clusters as well (main-sequence fitting). This technique is useful because it uses a large number of stars to
reduce statistical errors (as compared to spectroscopic parallax). One may use RR Lyraes in clusters of
known distance to determine intrinsic luminosities.
Secular parallax may also be used by repeating measurements over several years and taking advantage of the
Suns peculiar motion with respect to a group of stars.

The Flat Rotation Curve and Evidence of Dark Matter
Compiled data shows a relatively flat rotation speed curve for the MW beyond the Suns galactocentric radius,
here taken to be 8.5 kpc (although the error bars are wide):

Clemens graph (DP Clemens1985)
494
of rotation speed versus galactocentric radius,
data from multiple authors plus Clemens paper , assuming R0=8.5 Mpc, 0(R0)=220 km s
-1
.
The near-flatness of this curve of rotation speed beyond R 8 Mpc is surprising, as most of the luminosity of
the Galaxy is produced by matter residing inside R0. Similar observations by Vera Rubin (b. 1928) and
others have been made for other galaxies (such as for NGC 2998, NGC 4378, etc.) and support the existence
of dark matter. (IMA2 p. 914 919)
The empirical formula (also presented earlier)
/ +/

494
Clemens, DP, Massachusetts--Stony Brook Galactic Plane CO Survey: The Galactic Disk Rotation Curve,
ApJ, 295, 422 (1985).
of Navarro, Frenk and White was proposed in 1996.
495
It is applicable over a large range and tapers better at
large radius (near the edge of the halo). However, it is still not unbounded to infinity, and it is possible that
other galaxies may overlap and perhaps share the dark matter (IMA2 p. 219 and my inference).

Galactic Center

The galactic center is located in the Sagittarius constellation. The IAU 1985 agreed-upon distance to the
galactic center from the Sun (the solar Galactocentric distance R0) is 8.5 kpc, but more recent data and the
authors of IMA2 have taken 8 kpc as the best working estimate (thus about 26,100 ly). The Sun lies about 30
pc above the central galactic plane, and is rising so that it will attain a maximum z of 85 pc above the plane in
15 million years, at which time we will have a better view of the center!
This regions is best viewed with radio, microwave, and infrared, and also X-rays and gamma rays. The K
band of IR is useful, and helps to show that the mass density of stars rises toward the center as r
-1.8
down to
r = 0.1 pc to 1.0 pc. The frequent interactions of stars at these short distances leads to their being
thermalized to an isothermal state just as if they were in a Maxwellian distribution like a stellar gas.
However, velocities increase near the center in a way suggesting that considerably more mass in present in or
near the center.
Sgr A* and Its Neighbors
Schdel and Genzel et al 2003
496
have shown stars orbiting at close distances from the galactic center at Sgr
A*, particularly star S2, which has an orbital period of only 15.2 years, eccentricity 0.87, a semimajor axis of
about 1000 AU, perigalacticon of 1.8 x 10
13
m = 120 AU = 17 light-hours. The mass interior to S2 must be
about 3.7 x 10
6
M. The Sagittarius A complex is described in IMA2 p. 927-929. An unusual molecular
circumnuclear ring is a doughnut shaped structure with inner radius 2 pc, perhaps a SN remnant with
clumpy areas from previous shocks...
Sgr West includes Sgr A*, which appears to coincide with a source of X-rays. Sagittarius A* is a strong radio
point source of < 2 AU size. The J2000.0 celestial coordinates of the galactic center (specifically the suspected
supermassive black hole Sgr A*) are approximately
Sgr A* = 17h 45m 40s
Sgr A* = 29 00 27.9
which differ slightly from the previously accepted coordinates of the galactic center.
Sgr A* appears to be the location of a supermassive black hole of mass 3.7 x 10
6
M and a Schwarzschild
radius of 16 R. (The Schwarzschild radius defines the event horizon.) The UV luminosity is 10
7
L, the
effective temp 35,000 K. The observed accretion rate is
10
-3
to 10
-2
M yr
-1
, which could easily provide the observed luminosity. Sgr A* appears to be nearly at rest in
the center of the galaxyit is extremely resistant to movement by gravitational tugs. Strong winds or ejected
material are blowing away from Sgr A*. The region must have been violent in the recent past. X-ray flare-ups
are observed of 160 times baseline activity currently.
497
Gamma rays have been detected, including 511 keV
photons from positron annihilations and the 1.8 MeV emission of decay of the relatively short-lived isotope
26
Al to
26
Mg. The region is clearly a dynamic environment. (However, other galaxies such as Andromeda
exhibit much more active galactic nuclei.)

495
Navarro JF, Frenk CS, & White SDM, The Structure of Cold Dark Matter Halos, ApJ v.462, p.563
496
Eisenhauer F, et al. A Geometric Determination of the Distance to the Galactic Center Astrophysical
Journal Letters Volume 597 Number 2 2003 ApJ 597
497
Baganoff FK, et al Chandra X-ray spectroscopic imaging of Sagittarius A* and the central parsec of the
Galaxy, ApJ 591:891915, July 2003, http://docs.lib.purdue.edu/physics articles/368

S2 is a star in a close elliptical orbit about the compact radio source Sgr A* marked by
498

498
Eisenhauer F, et al., 2003, op. cit.
The Nature of Galaxies (Chapter 25)

Definition and Scope of Galaxies
A galaxy is a massive, gravitationally bound system that consists of stars and stellar remnants, an interstellar
medium of gas and dust, and a poorly understood component called dark matter. Many galaxies, particularly
bright ellipticals and spiral galaxies have supermassive black holes at their centers. Sizes of galaxies range
from dwarfs with as few as 10
7
stars to giants with as many as 10
14
stars, each star orbiting the galaxy's
center of mass. There are probably more than 170 billion galaxies in the observable universe.
499

Early Galaxy Discoveries
After early ideas and explorations regarding galaxies of Kant and mapping of the MW by Herschel, systematic
cataloging of the galaxies was done by Charles Messier (17301817). The first version of Messier's Catalogue
des Nbuleuses & des amas d'toiles contained 45 objects and was published in 1774. The final version of
the catalogue was published in 1781the final list of Messier objects had grown to 103.
500
On several
different occasions between 1921 and 1966, astronomers and historians discovered evidence of another seven
deep-sky objects that were observed either by Messier or his friend and assistant, Pierre Mchain, shortly
after the final version was published. These seven objects, M104 through M110, are accepted by astronomers
as "official" Messier objects.
501
They include gaseous nebulae (including supernova remnants, such as the
Crab M1, and giant molecular clouds and H II regions like M16 Eagle and M42 Orion Nebula), planetary
nebulae (M27, M57, M76), and globular and open stellar clusters, asterisms (M73), as well as galaxies.
Additional catalogs were made by William Herschel (17381822) and his son John. J. L. E. Dreyer (1852
1926) first published the New General Catalogue of Nebulae and Clusters of Star (NGC) of 7,840 objects in
1888most of the Messier objects were included.
William Parsons (18001867) first observed spiral structure in galaxies, and Doppler shift confirming
rotations was demonstrated first by Vesto M Slipher (18751969).
There was great uncertainty about the true nature of many nebulas, culminating in the Great Debate in 1920
of Harlow Shapley (who thought galaxies were local) and Heber D. Curtis (who thought galaxies were
extragalactic). Both were drawing on flawed data. Edwin Hubble (18891953) settled the debate when he
resolved M31 and M33 into individual stars (1923-4) and proved with Cepheid variables that Andromeda M31
was 285 kpc distant (now thought to be 770 kpc), and a galaxy like our own.
The realization that a myriad of galaxies exist other than our own has been of importance comparable to the
Copernican Revolution. (IMA2 p. 942)
Fritz Zwicky (1898-1974) first observed in 1933, using the virial theorem, that there is not nearly enough
luminous matter to account for observed orbits of cluster galaxies such as the Coma galaxy cluster, and
posited unseen dunkle Materie (dark matter).
502

Classification of Galaxies
This is reviewed in IMA2 p. 942948 and only selected comments are included here. Contributions were made
by Edwin Hubble, Sidney van den Bergh, Gerard de Vaucouleurs (19181995), and the Yerkes (or Morgan)
scheme of William Morgan and Philip Keenan.
Hubble introduced in 1926 a graduated sequence (the tuning fork diagram) extending (incorrectly) from
early to late types. The earlier types were ellipticals E0, E1, E2, E3, E4, E5, E6, E7 (which were divided
based on numerical degree of apparent ellipticity 10 = 07, where eccentricity 10 (1 (b/a)). The
transitional lenticulars S0/SB0 could be further subdivided into S01/SB01, SO2/SB02, and S03/SB03.

499
partially paraphrased from http://en.wikipedia.org/wiki/Galaxy
500
http://en.wikipedia.org/wiki/List_of_Messier_objects
501
http://en.wikipedia.org/wiki/Charles_Messier
502
http://en.wikipedia.org/wiki/Fritz_Zwicky#Dark_matter
The later types were spirals (Sa, Sb, Sc, SBa, SBb, SBc, depending on the size of the nucleus and the
tightness of the spiral arms) and Irregulars Ir. The B in spirals indicated barred (like our Milky Way: SBc).
Galaxies were further evaluated for oblateness (squashed top to bottom) and prolateness (elongated like a
football). In reality, galaxies may have 3 orthogonal axes of unequal length and the inclination can be hard to
determine, complicating the analysis.
de Vaucoulers introduced variations in classifying galaxies, adding Sd, SBd, Sm, SBm, Im and Ir, and called
normal spiral SA, weak bars SAB, and added recognition of dust S01, S02, S03.
The modern sequence is a confused hybrid: E0, E1, E2, E3, E4, E5, E6, E7, S01, S02, S03, Sa, Sab, Sb, Sbc,
Sc, Scd, Sd, Sm, Im, and Ir. A similar sequence with Bars B: SB01, SB02, SB03, SBa, SBab, SBb, SBbc,
SBc, SBcd, SBd, SBm. Other taxonomists have added R for outer rings, and s or rs or r for no inner ring,
weak inner ring, or strong inner ring, respectively.
There is also a cD-type galaxy. The "c" in "cD" refers to the fact that the galaxies are very large, hence
supergiant, while the "D" refers to the fact that the galaxies appear diffuse. cD's are the largest galaxies and
may result from galaxy mergers in galaxy clusters.
503

Sidney van den Bergh introduced luminosity classes with Roman numerals I V, and in-between
designations are made. For example, the Milky Way is fully classified SBbcIII.
Modern surveys like SDSS utilize automated galaxy classification.
504

Spiral and Irregular Galaxies

Spiral galaxies each have (1) a disk of stars, gas, and dust rotating about the center, (2) a central stellar bulge
resembling an elliptical galaxy, (3) a nearly spherical halo of stars, many in globular clusters, and (4) a
supermassive black hole at the very center.
505

Galaxy brightness may be affected by extinction.
A K-correction must be used for distant extragalactic objects due to redshift. (IMA2 p. 950) The need for a
K-correction arises because an astronomical measurement through a single filter or a single bandpass only
sees a fraction of the total spectrum, redshifted into the frame of the observer. So if the observer wants to
compare the measurements through a red filter of objects at different redshifts, the observer will have to apply
estimates of the K corrections to these measurements to make a comparison...
506

The brightness of the background sky must also be considered and subtracted from measurements on
galaxies. Like other extended objects, the sky brightness sky is expressed as surface brightness, usually in
units of magnitude per square parsec. The darkest sky brightness
507
is about 22 mag/arcsec
2
or 220 S10
units(where an S10 unit is defined as the surface brightness of a star whose V-band magnitude is 10 and
whose light is smeared over one square degree, or equivalently 27.78 mag arcsec
-2
). Other sites have V-band
sky brightness in mag./arcsec
2
as follows: Mount Wilson 19.8, Palomar 21.5, Haute Provence 21.8, Lowell
Mars Hill 20.5, Atacama Paranal
508
site 22.6 (in blue). Modern CCDs can measure brightness levels as low as
29 mag arcsec
-2
. (IMA2 p. 950)

503
cD-type galaxies:
http://en.wikipedia.org/wiki/Type-cD_galaxy and
Tonry JL, Properties of cD galaxies, Structure and dynamics of elliptical galaxies. Reidel Publishing Co.,
1987, p. 89-96
504
http://www.astro.virginia.edu/class/whittle/astr553/Topic02/Lecture_2.html
505
http://en.wikipedia.org/wiki/Spiral_galaxy
506
http://en.wikipedia.org/wiki/K_correction
507
Sky Brightness:
http://en.wikipedia.org/wiki/Sky_brightness
http://www.astropix.com/HTML/L_STORY/SKYBRITE.HTM
508
http://www.eso.org/~fpatat/science/skybright/zenit/zenit_paper.htm

World Atlas of the Artificial Night Sky Brightness
509

Isophotes are contours of constant surface brightness. These are used to define the radii of galaxies, since
there is no clearly defined cutoff. The Holmberg radius rH is a commonly used standard, being the projected
length of the semimajor axis of an ellipsoid having an isophotal surface brightness of
H = 26.5 B-mag. arcsec
-2
(i.e., in blue light). Clearly to use a standard this faint in terrestrial astronomy, one
must subtract out the sky brightness.
The effective radius re is defined as the projected radius within which one-half of the galaxys light is emitted.
The quantity e = surface brightness at the effective radius re.

de Vaucouleurs and Srsic (Sersic) Profiles
A purely empirical relationship between surface brightness for arbitrary r, namely (r), is given for spiral
galaxy bulges and for large ellipticals by an r
1/4
or de Vaucouleurs profile previously mentioned in units of
L per pc
-2
,

but here written in units of mag arcsec
-2
(band not stated):

+ [(
)
/
]
In contrast, disks including spiral disks are frequently modeled in surface brightness units of mag arcsec
-2
by
the empirical relationship

+ [(
)
/
]
where hr is the characteristic scale length along the midplane and the Sersic index = 1 (shown here as 1/1), a
relationship called exponential. In general, the exponent takes on the form of 1/n for differing Srsic
Profiles, where n = Srsic index and controls the degree of curvature of the profile (larger n makes the
surface curve more gradually over greater distance). Most galaxies are fit by Srsic profiles with indices in the
range 1/2 < n < 10. The best-fit value of n correlates with galaxy size and luminosity, such that bigger and
brighter galaxies tend to be best fit with larger n.
510

509
http://www.lightpollution.it/worldatlas/pages/fig1.htm
510
http://en.wikipedia.org/wiki/Sersic_profile
Rotation Curves of Galaxies
The Sersic profiles alone do not convey information about the distribution of dark matterthis may be
gleaned from the galaxys rotation curve and by other means.
Vera Rubin studied these in the 1970s and reported in 1975 her discovery of the apparently non-Newtonian
velocities of stars in spiral galaxies. (See also Differential Galaxy Rotation pertaining to the Milky Way,
above.) With her colleagues, she published families of rotation curves for spiral galaxies of types Sa, Sb, and
Sc in 1985. These compared the rotational (tangential) linear velocity to the radius, extending from the
galactic center to the R25 isophotal radius. The R25 is the isophotal radius for which blue-band surface
brightness has fallen to 25 B-mag/arcsec
2
.
511
The flattening of the curves past about 0.25 R25 to 0.5 R25 is
apparent. (The comparable diameter, D25 = 2R25, is also used in expressing galaxy size.)

For a particular Hubble galaxy type, galaxies that are more luminousi.e., have more negative MBhave
more rapidly rising and larger final values of Vmax. For the same luminosity, early Sa galaxies have larger
Vmax than late Sc galaxies. In general, the rotation curves suggest that the gravitational potentials are not
following the distribution of known luminous matter, thought to be a signature for dark matter (IMA2 p. 951)
The Vmax rotation speed of irregular galaxies is only 50 to 70 km s
-1
, perhaps indicating a lower threshold limit
on Vmax below which well-formed spiral galaxies do not develop.
See also discussion under Ellipticals: Shape and Rotation regarding rotationally supported shapes versus
pressure-supported shapes.

511
Rubin VC, et al, Rotation velocities of 16 SA galaxies and a comparison of Sa, Sb, and SC rotation
properties, ApJ, Part 1, vol. 289, Feb. 1, 1985, p. 81-98, p. 101-104.
The Tully Fisher Relation
These empirical formulas express a roughly linear relationship between luminosities of spiral galaxies in
various wavelength bands versus their Vmax. It was developed by R. Brent Tully and J. Richard Fisher and
published in 1977. It has been applied to 21-cm radio H I emissions (which exhibit double peaks from the
Doppler shifted galactic periphery approaching and receding, shown below), blue-band visible light MB (good
for young hot stars), and infrared wavelengths such as MH
i
(good for overall distributions of luminous mass).
(IMA2 p. 952-6). Distance measurement discussed later.

21-cm radio H I emissions from galaxy NGC 3198, which exhibit double peaks
from the Doppler shifted galactic periphery segments approaching toward
and receding from the observer. Spectral Flux density in Janskys.
VHEL is heliocentric (Sun-based) velocity in km s
-1
,

averaging 660 km s
-1
recession(Begeman 1989).
512

The Radius-Luminosity Relation
For early spiral Sa-Sc galaxies of increasing luminosity, their isophotal R25 radii (measured in kpc) increase
according to an empirical relationship. (IMA2 p. 956):
log10 R25 = -0.249MB - 4.00

512
Begeman KG, H I rotation curves of spiral galaxies. I - NGC 3198, Astronomy and Astrophysics, vol. 223,
no. 1-2, Oct. 1989, p. 47-60

Mass and Mass to Light Ratios
Scs are slightly less massive than Sas. However, the average ratios of M/LB (expressed in units of
)
are more strongly dependent on Hubble type, being M/LB 6.2 for Sas, 4.5 for Sbs , and 2.6 for Scs.

Color, Gas, and Dust
The trend of Mass to Light Ratios suggests a greater fraction of massive main sequence early stars in late
Scs compared to early Sas (not inconsistent and unfortunate Hubble labels), and therefore a bluer color
index. Scs also have greater gas and dust and are more centrally condensed.
Irregulars (such as the LMC) are the bluest of galaxies, suggesting active star formation.
Stars in spiral galaxies are redder in their bulges than in their disks, which relates to metallicity and the
relative proportion of star formation. (IMA2 p. 957-8)

X-ray Luminosity
There is a close linear correlation between X-ray and B-band luminosity, suggesting that X-rays are generated
by a small but constant fraction of objects (such as SNRs) in a spiral galaxy.

Supermassive Black Holes SMBH
There is strong evidence of a SMBH at the center of Andromeda M31, where a large non-luminous mass is
predicted with mass = 1.4x10
8
M. Another SMBH with a mass of 3.2x10
9
M has been inferred in M87, a
giant elliptical galaxy.

Black holes mass can also be deduced by velocity dispersion via the virial theorem (IMA2 p. 959-60), using the
derived formula

where the sigma term is the velocity dispersion in the radial velocity and R is the radius of a suitably chosen
sphere within the black holes influence.
Another empirical relation with dispersion is given by Mbh = (/0)
.
It is quite possible that a SMBH is at the heart of most galaxies.

Globular Star Clusters
To compare the frequency of globular clusters (GCs) in galaxies, the quantity specific frequency SN of GCs is
used. This term that has been normalized to a standard absolute magnitude of the parent galaxy. SN is
greater in late-type galaxies and in galaxies with higher luminosity. In particular, elliptical galaxies and
especially the gigantic cDs have more GCs per unit luminosity.

Specific frequency of globular clusters,
evaluated by luminosity and galaxy Hubble type (Harris 1991
513
)

Spiral Structure
There is a rich and elegant body of mathematical theory which I will report only selectively here. Spiral
galaxies may have a grand design (such as M51 shown at the beginning of this document) with two very
symmetric and well-defined arms, or may have multiple arms, or at the other extreme may be poorly defined
flocculent spirals with only partial arm fragments.

513
Harris WE, Globular cluster systems in galaxies beyond the Local Group, Annual review of astronomy and
astrophysics. Vol. 29 (A92-18081 05-90,. 1991
The arms are especially apparent and most distinct in blue band, due to the presence of hot young O and B
blue stars, consistent with active star formation. In infrared or red light, the arms appear broader and less
pronounced, suggesting that the disk is dominated by older redder stars. Dust and gas preferentially resides
on the concave sides of the arms. In general, the spiral bands appear quite different when viewed in varying
wavelengths, such as X-ray, UV, IR, and radio.
The overall [?number] density of stars is increased only by about 10 to 20% in spiral arms (ZXI lecture 4).
The convex side of the arm might be expected to be the forward edge (the tip curves back behind, pointing in a
direction essentially opposite to the direction of rotation). The concave inner side is the trailing edge. This
type of rotation is called a trailing arm structure. However, this is not always the case, the direction of
rotation can be hard to determine, and some spiral arms are leading structures. Spiral arms that can be
evaluated are usually trailing, however.
Spiral arms winding might be expected if they were material structures composed of a fixed set of
differentially rotating stars, but spiral arms probably do not wind up.
Instead, the best hypothesis explaining spiral arms is the Lin-Shu Density wave theory (of C. C. Lin and
Frank Shu, devised in 1964.
514
) The arms are quasi-static density waves (which are similar to a local traffic
piling up on a freeway, in which individual cars slow and accumulate in the congested area while passing
through). In a non-inertial rotating frame with angular global patterns speed gp, they appear to be
stationary. Stars, dust, and gas move through these regions of higher density (perhaps 10% or 20% higher in
density). At the corotation radius RC, the spiral arm and the objects at the same radius move together.
Objects closer to the center (R < RC) rotate in the non-inertial frame with shorter periods than the arm,
whereas objects further away (R > RC) rotate more slowly than the density wave pattern. Most new star
formation occurs closer in to the center than RC. Density wave formation can cause a 3 or 4 armed spiral,
and there are numerous other variants.

A pileup of nested ellipses rotating at slightly different angular velocities
may explain a density wave having the shape of spiral arms with a central bar.
515

It is also possible to have mathematically defined resonances leading to denser regions where stars
accumulate:
0 to 2 inner Lindblad resonances (ILR), which may cause inner rings,
an ultraharmonic resonance, which may cause an inner ring,
a corotation resonance (CR), which may bound bars, and
an outer Lindblad resonance (OLR), which may cause outer rings.
Individual stars rotating about the galactic midplane exhibit motion that to a first order approximation can
be imagined as being the combination of a retrograde orbit about an epicycle and the prograde orbit of the
center of the epicycle about a perfect circle (IMA2 p. 975).
A companion galaxy may be needed to cause the initial tidal interactions triggering spiral arm formation.

514
http://en.wikipedia.org/wiki/Density_wave_theory
515
ibid.
Impressive spiral arms and bars have been demonstrated in N-Body simulations.
516

In the case of flocculent spirals, an alternate theory of stochastic self-propagating star formation has been
advanced by Mueller and Arnett, 1976. The theory posits outbursts of star formation that propagate across
the galaxy.
The number and how tightly the spiral arms are wound are well correlated with other, large scale properties
of the galaxies, such as the luminosity of the bulge relative to the disk and the amount of gas in the galaxy.
This suggests that there are global physical processes involved in spiral arms. (ZXI lecture 3).

Elliptical and SO Galaxies

These are remarkably diverse and complex. An elliptical galaxy is a galaxy having an approximately
ellipsoidal shape and a smooth, nearly featureless brightness profile. They range in shape from nearly
spherical to highly flat and in size from hundreds of millions to over one trillion stars.
517

They tend to have older (and thus reddish) lower-mass stars, with a majority having sparse ISM and dust and
minimal star formation activity, with few open star clusters. Unlike spirals, many are dominated not by
rotation but by random stellar motions or pressure. Rather than Hubbles system of categorizing ellipticals
based on ellipticity (0 - 7), it has been found more useful to categorize them by other factors such as size,
absolute magnitude, and surface brightness (IMA2 p. 983).

516
Hohl F, Numerical Experiments with a Disk of Stars ApJ 168 p. 343, 1971
517
http://en.wikipedia.org/wiki/Elliptical_galaxy , edited

Lenticular S0 Galaxy NGC 4150, mass~ 6x10
9
M, distance ~13.6 Mpc, MB = -18.5 mag
The inset core width is ~500 pc. It shows a slightly rotated magnified view of the chaotic activity
inside the galaxy's core. The blue areas indicate a flurry of recent star birth.
The stellar breeding ground is about 1,300 light-years across. The stars in this area
are less than a billion years old. Dark strands of dust are also apparent, tentative evidence
of a recent merger. (HST images, Oct. and Nov. 2009)
518

Morphological Classes of Elliptical and SO Galaxies
They include:
cD Galaxies:
cDs are immense, rare, up to 1 Mpc in size, found near center of large dense galactic clusters. They
have MB magnitudes of 22 to 25, masses of 10
13
to 10
14
M, and D25 diameters of 300-1000 kpc.
They have high central surface brightness ( = 18 B-mag arcsec
-2
) and very extended diffuse envelopes
(for which = 26 27 B-mag arcsec
-2
). They have large numbers of globular clusters and a high
specific frequency for globular clusters (averaging SN =15). They have high mass to luminosity ratios
(M/LB >100 M/L, as high as 750 M/L), indicating large amounts of dark matter. Surface
brightness profiles follow the r
1/4
law.

518
Lenticular S0 Galaxy NGC 4150:
http://hubblesite.org/newscenter/archive/releases/galaxy/elliptical/2010/38/
http://hubblesite.org/pubinfo/pdf/2010/38/pdf.pdf
Normal Elliptical Galaxies:
These include giant ellipticals gEs, intermediate luminosity ellipticals Es, and compact ellipticals
cEs. Es have MB magnitudes of -15 to -23, masses of 10
8
to 10
13
M, and D25 diameters of 1-200
kpc. They have relatively high central surface brightness . They have specific frequency for globular
clusters SN =5. They have high mass to luminosity ratios (M/LB = 7 to 100 M/L). Hubbles
Lenticular S0s and SB0s are often grouped here as well, and have MB magnitudes of -17 to -22,
masses of 10
10
to 10
12
M, and D25 diameters of 10-100 kpc, M/LB ~10 M/L, specific frequency for
globular clusters SN ~ 5. Surface brightness profiles follow the r
1/4
law (such as shown on IMA2 p. 986
for NGC 3379)

Dwarf Elliptical Galaxies dEs:
dEs have MB magnitudes of -13 to -19, masses of 10
7
to 10
9
M, and D25 diameters of 1-10 kpc. They
have lower surface brightness than cEs of the same magnitude. They have specific frequency for
globular clusters SN ~ 4.8, still higher than spirals. They have mass to luminosity ratios M/LB = ~10
M/L Metallicities are lower than for Es. Surface brightness profiles are exponentialthey follow
the r
1/1
law, has a linear relation to r/hr, and I(R) e
-kR
.

Dwarf Spheroidal Galaxies dSphs
dSphs are extremely low luminosity low surface brightness objects detected only in the vicinity of the
MW. They have MB magnitudes of -8 to -15, masses of 10
7
to 10
8
M, and D25 diameters of 0.1-0.5
kpc. They have mass to luminosity ratios M/LB = 5 100 M/L. No clusters are mentioned. Surface
brightness profiles are exponential.

Blue Compact Dwarf Galaxies BCDs:
BCDs are small compact galaxies that are unusually blue, with color index B - V = 0.0 to 0.3, thus
like spectral class A stars, suggesting they are undergoing active star formation. They have MB
magnitudes of -14 to -17, masses of ~10
9
M, and D25 diameters of < 3 kpc. They have low mass to
luminosity ratios M/LB = 0.1 10 M/L They have an abundance of H I and H II. No clusters are
mentioned.

Gravitational Binding, Dust and Gas, and Star Formation
Ellipticals (dEs and dSphs) tend to have low gravitational binding and retain only small amounts of gas and
dust (esp. dSphs). Therefore they are not now actively forming stars. The dust sometimes counterrotates
with respect to the stellar components. Perhaps gas has been stripped by supernovas or galaxy interactions.
What interstellar medium is present tends to be hot, with T > 10
6
K (ZXI lecture 5).

Metallicity
Ellipticals have low metallicities, such as seen with globular clusters. Metallicity correlates with luminosity
metallicity of dwarf elliptical and ellipticals tends to increase with absolute blue-magnitude MB (IMA2 p. 958).
As with spirals, metallicity gradients are seen with ellipticals...

The Faber-Jackson Relation
This expresses, for dEs, dSphs, and Es as well as for spiral bulges, a correlation between central radial
velocity dispersion 0 and the luminosity corresponding to MB, namely

where was originally taken to be 4 by Faber and Jackson,
519
but is now fitted to a range of values. IMA2 p.
988 describes further efforts to fit the observations to an empirical equation relating luminosity, velocity
dispersion, and effective radius re. The relation exhibits considerable scatter for ellipticals.

Shape and Rotation
The source of the shape of ellipticals is somewhat unclear. Most have no preferred axis of rotation. It is likely
that some of the gas, dust, globular clusters, and even dwarf galaxies have been captured since the galaxies
formation. Their shapes are not generally due to rotation, as the velocity dispersions often exceed the
rotational velocities. The shapes may be due to anisotropic velocity dispersions.
As an extreme example, the E3-E4 galaxy NGC 1600 has Vrot = 1.9 km s
-1
and an extremely low Vrot / =
0.013.
The rotation parameter (V/)* is defined as
(

Here, the isotropic term in the denominator is the value expected for an isotropic oblate spheroid flattened
purely by rotation.
520
A galaxy is considered to be rotationally supported if rotation parameter (V/)* 0.7.
However, it is < 0.016 for this galaxy, thus it is not rotationally supported. NGC 1600 has no statistically
significant axis of rotation (IMA2 p. 989).
Most bright Es and gEs have (V/)* ~ 0.4, and are considered to be pressure-supported, meaning that their
shapes are due to random stellar motions as with an ideal gas, rather than to overall rotation.
However, less luminous galaxies for which -18 > MB > -20.5 (including cEs) typically have rotation parameters
(V/)* 0.9 and are therefore rotationally supported. (IMA2 p. 989) (Spiral bulges are also generally
rotationally supported.)
The shapes of these elliptical galaxy isophotal surfaces are subtly boxy or disky. In the Fourier expansion
of the polar coordinates representation of shape, given by radius a() of this surface, the term a4 cos(4) is
associated with the boxy/disky shape component. If a4 < 0, the surface is boxy (and these tend to be
pressure supported, etc.), whereas if a4 > 0, the surface is disky (and these tend to be rotationally supported,
etc.) Although there are many factors involved, boxy galaxies may reflect a signature of past galaxy mergers
or tidal interactions (IMA2 p. 991).

Relative Numbers of Galaxies by Hubble Type

In order to express the relative number of galaxies by Hubble type, it is useful to define the luminosity
function LF expressed as (M) dM, which gives the number of galaxies in a sample with absolute magnitudes
between M and M+dM. (M) has units of galaxies per magnitude per pc
3
. The following graph (Binggeli,
Sandage, Tammann 1988)
521
gives the log of luminosity functions (M) for a selection of 121 galaxies near the
MW (local field) and also for a much larger sample of 1300 galaxies in the Virgo cluster (which is considered
much more reliable). MBT is absolute blue-band total magnitude (i.e., not surface brightness), and Total is
the sum of the individual luminosity functions (including types that are not specifically graphed; see also
IMA2 p. 991). Galaxies of type dE and dwarf Irr are the greatest in numbers in Virgo, but Ss and Es are the
most prominent in cluster luminosity and mass. [The latter statement in IMA2 is not quite so obvious to me,
but presumably true due to the much higher magnitudes of the S and E galaxies]. The ellipticals are more

519
Faber SM, Jackson RE, Velocity dispersions and mass-to-light ratios for elliptical galaxies, Astrophysical
Journal, vol. 204, Mar. 15, 1976, pt. 1, p. 668-683
520
Houjun Mo, Frank Van Den Bosch, Simon D. M. White, Galaxy Formation and Evolution, Cambridge UP,
2010, p. 47.
521
Binggeli B; Sandage A, Tammann GA, The luminosity function of galaxies, Annual review of astronomy
and astrophysics. Vol. 26, 1988, p. 509-560.
prominent in Virgo than the local field, and BCDs are only seen in Virgo. The Virgo values reported (Binggeli
p. 550) are Es=12%, SOs=26% and S+Irrs=62%, whereas the values for the Coma cluster (same page) differ
significantly. Clearly the relative proportions of galaxy types depends on the population sampled.

Graph of Binggeli, Sandage, Tammann 1988.
522

A luminosity function (M,z) can also be derived, a function of both luminosity and redshift distance z, thus
in the luminosity-position plane.

522
ibid.
ZXI (lecture 6) showed a graph from an article he co-authored in which blue galaxies have a different LF from
red galaxies.
523

The following graph of SDSS data plots a luminosity function (M,z) against Mr for z=0.1, which is near the
median redshift of the sample. I have not mastered this complex subject matter. The authors measured the
galaxy luminosity density at z = 0.1 in five optical bandpasses corresponding to the SDSS bandpasses shifted
to match their rest-frame shape at z = 0.1. These included
0.1
r, for which Teff = 5595 K for the r bandpass. In
this case, is not decomposable into a product of separate distributions f(M) and n(z). The relative peaks on
the left found in small numbers represent the largest ellipticals, including cD galaxies.
524

523
Baldry IK et al, Quantifying the Bimodal Color-Magnitude Distribution of Galaxies, The Astrophysical
Journal, Volume 600, Issue 2, pp. 681-694, 01/2004. Figure 7.
524
Blanton MR et al, The Galaxy Luminosity Function and Luminosity Density at Redshift z = 0.1, The
Astrophysical Journal Volume 592 Number 2 2003.
Galactic Evolution (Chapter 26)

Interaction of Galaxies

Most galaxies have not arisen in isolation but in groups or clusters (IMA2 p. 999), and their interactions play
an important role in their evolution. Intergalactic spacing is typically only 100 times their diameters, and
some clusters (such as Coma Cluster) are more densely populated. Ellipticals are more common in the center
areas than the outer regions of such galaxy clusters, especially for well-formed clusters. These observations
may correlate with an increased probability of galaxy interactions and mergers, leading to increased velocity
dispersion and a r
1/4
de Vaucouleurs profile.
Many disk galaxies have warped disks in H I emission.
The gas in galaxies in a cluster appears to be pulled out of the galaxies into the surrounding intergalactic
space in the cluster, where it is seen to be hot and X-ray emitting.

Dynamical Friction
The individual stars in colliding galaxies rarely physically collide, due to their small size relative to the
interstellar spacing, but galaxies themselves do commonly collide or pass through each other. Dynamical
friction describes the loss of momentum and kinetic energy of moving bodies arising through an indirect
gravitational interaction with the surrounding matter in space. It is sometimes referred to as gravitational
drag, and was first discussed in detail by Chandrasekhar in 1943.
525

This quasi-frictional force takes the form of

where in an idealized example
M is the mass of a smallish approaching object Obj1 (perhaps a globular cluster or small galaxy)
vM is the average velocity of Obj1 relative to Obj2,
is the density of Obj2, where Obj2 is a very extensive object (such as a large galaxy),
C is a function that depends on how vM compares to the velocity distribution of Obj2, and
fd is the resultant drag force on Obj1.
A wake trail is left behind Obj1 as it passes through Obj2, and some of Obj1s kinetic energy is transferred to
the stars of Obj2. (MCM: I am assuming that the stars in this trail have arisen from both Obj1 and Obj2)
The frictional force is stronger for greater . It is also apparent that slow encounters with small vM exert
much more dynamical frictional drag on Obj1 than fast fly-throughs, for which there is less time for
interaction.
The dynamical friction force exerts a torque on clusters, reducing their angular momentum, causing many of
them eventually to spiral in to the center of their galaxy and disappear, thus depleting the galaxy of star
clusters, especially the lower mass clusters. (Details of the explanation are somewhat unclear to me, see
IMA2 p. 1001-4 and here
526
). Even satellite galaxies are affected, explaining why it is believed that the
Magellanic clouds will someday merge with the MW.
Frictional effects from a collision with a galaxy in a retrograde orbit may explain why a galaxy can end up with
a counterrotating core.

525
http://en.wikipedia.org/wiki/Dynamical_friction
526
http://www.research-in-germany.de/93492/2012-02-15-how-globular-star-clusters-survive-
collisions.html
Rapid Encounters, Rings, Tidal stripping, and Streams
Collisions at fast vMs cause an increase of KE in the collided galaxies. In order for the galaxies to return to
virial equilibrium, the added KE must be adjusted for, either by increasing star spacing, or by having it
carried away from the galaxy, often by the most energetic components in the form of a stream of gas and
stars (cf. the Magellanic Stream). This evaporation cools the galaxy. [the Magellanic stream began] at
about the time when the two Magellanic Clouds may have passed close to each other, triggering massive
bursts of star formation.... The strong stellar winds and supernova explosions from that burst of star
formation could have blown out the gas and started it flowing toward the Milky Way
527

High speed nearly head-on galactic collisions may also result in a ring galaxy such as the Cartwheel (ESO
350-40), in which the ring represents a sudden brief burst of star formation:

Cartwheel Galaxy (ESO 350-40), composite image in false colors, using
(1) Galaxy Evolution Explorer in ultraviolet light (blue); (2) HST in visible light (green);
(3) the Spitzer Space Telescope in IR (red); and (4) Chandra X-ray Observatory (purple).
The Cartwheel galaxy is one of the brightest ultraviolet energy sources in the local universe.
In some visible-light images, it appears to have spokes.
Distance 500 Mly = 150 Mpc, ~1 arcmin (NASA etc.)
528

Gaseous rings may be formed by collisionsthere are no stars in these rings (see for instance M105).
Galaxies that are bound to each other eventually merge.

527
http://www.astronomynow.com/news/n1001/05streamer/
528
Cartwheel Galaxy (ESO 350-40):
http://www.nasa.gov/centers/jpl/news/galex-20060111.html
http://en.wikipedia.org/wiki/Cartwheel_Galaxy
Tidal stripping can dissipate the orbital energy by
pulling stars and gas out of one or both galaxies,
such as caused the Magellanic Stream. This
occurs when material extends beyond (overflows
or escapes from the potential well of) the retaining
equipotential lines. These potential barriers are
analogous to Roche lobes for stars.
The diagram
529
to the right depicts a gravitational
equipotential surface, showing potential wells
bounded by Roche Lobes (not explicitly delineated)
that are separated by the L1 Lagrangian point,
here for close binary stars and by analogy for
galaxies. (See full discussion of a similar diagram
in Chapter 18)
The tidal radii l1 and l2 define the respective
distances to the L1 Lagrangian point (the point
where escape to the other object is the easiest, see
diagram below) for the two colliding objects. (IMA2 p. 1006, 657, 659) Tidal stripping may also explain the
retrograde orbits of about one half of the stars in the outer stellar halo of the Milky Way (IMA2 p. 1009)
Collisions may also result in tidal tails, such as are dramatically evident in the well-known colliding
Antennae galaxies NGC 4038 and 4039.
530
According to simulations, tails are usually seen only with close
slow encounters, which produce tidal deformations in the equipotential surfaces analogous to Earths tidal
bulging. The stars and gas torn from the near side of a tidally bulging colliding galaxy form an apparent
bridge to the other galaxy (often obscured, may not actually connect in an optimal view), whereas the material
bulging and escaping on the opposite side will form a tail that conserves angular momentum (IMA2 p. 1009
and here
531
).
Galaxies have been found that have polar rings
532
(usually around ellipticals or SO galaxies, such as NGC
4650A) or are dust-lane ellipticals (such as NGC 5128), in either case exhibiting material stripped from
colliding galaxies.

Modeling Interactions With N-Body Simulations
This approach has become an indispensable but complex and costly component of exploring galactic
interactions. It was pioneered by Alar and Juri Toomre in 1972. Modern simulations must include dark
matter and self-gravitation. Examples of simulations are available on the Web,
533
including links at the
beginning of this document. Simulations have helped for instance to show why outer clusters of the Milky
Way were probably formed elsewhere and captured later . Satellite galaxy collisions may explain the warps
that persist for billions of year in some galaxy disks. (Details are mostly omitted here. See IMA2 p. 1008-
1011, etc.)

Starburst Galaxies
These are often colliding and interacting galaxies emitting blue color from prominent hot new OB star
formation. They have increased luminosity (often obscured by dust in visible and UV bands, but bright and
well seen in IR). The star formation may be very rapid, using up to 300 M per year, but may only last 20
million years. Star formation is not necessarily concentrated in the galactic center, and may extend over
much of the disk. M82 is an example:

529
http://csep10.phys.utk.edu/astr162/lect/binaries/accreting.html
530
http://en.wikipedia.org/wiki/Antennae_Galaxies
531
http://en.wikipedia.org/wiki/Galactic_tide
532
http://en.wikipedia.org/wiki/Polar-ring_galaxy
533
http://hubblesite.org/newscenter/archive/releases/2008/16/video/d/ etc.

M82 (NGC 3034) Starburst Galaxy by HST, 2006
Galaxy at distance 3.7 Mpc, width 8 arcmin or ~8.5 kpc (thus smaller than the MW)
Composite of Multiple filters: F658N (Hydrogen-alpha) assigned red/orange; F814W (I) assigned red
F555W (V) assigned green; and F435W (B) assigned blue
534

According to some definitions, starburst galaxies include Blue compact galaxies (BCGs), Luminous infrared
galaxies (LIRGs) and Ultra-Luminous infrared galaxies (ULIRGs), and Wolf-Rayet galaxies. Irregular galaxies
often exhibit knots of starburst, often spread throughout the irregular galaxy... To ignite a starburst, it is
necessary to concentrate a large amount of cool molecular gas in a small volume. Such concentrations and
perturbations are strongly suspected to cause global starburst phenomena in major galaxy mergers, although
the exact mechanisms are not fully understood.
535

Mergers in Elliptical and cD Galaxies
cD elliptical galaxies are probably the product of mergers. Most have multiple nuclei, and cDs are often
located at the bottom of a clusters potential well. Normal ellipticals also show evidence of mergers, especially
boxy Es.
A majority of ellipticals (such as NGC 3923) have shells of stars partially enclosing them, which may be due
to captured stars (from small faint galaxies) sloshing back in forth in their potential wells (IMA2 p. 1014).
Many and perhaps all large elliptical galaxies are strongly influenced by mergers. (IMA2 p. 1015) The r
1/4

profiles are normal outcomes of mergers. However, dwarf ellipticals and dwarf spheroidals do not result from
mergers.

534
535
http://en.wikipedia.org/wiki/Starburst_galaxy
Binary Supermassive Black Holes
Galaxies, each having a black hole, that merge can lead to binary black holes orbiting each other, for instance
probably NGC 6240 which appears to have double nuclei orbiting at about 0.8 kpc.
536
Such mergers may be a
pathway to the growth of supermassive black holes (IMA2 p. 1016).

The Formation of Galaxies

A limited summary follows.

Eggen, Lynden-Bell, and Sandage ELS Collapse Model of the Milky Way
An early attempt to model our Milky Way Galaxy formation, 1962, proposed a top-down approach that
involves the differentiation of a single, immense proto-Galactic cloud. But this is contradicted by 1/2 of outer
halo stars having retrograde orbits. Another problem is the disparity in apparent ages among globular
clusters (and other problems are mentioned).
Galaxy formation theory is not mature, and remains incomplete. We need a stellar birthrate function, based
on star formation rate SFR and the initial mass function IMF. ...
The G-dwarf problem
This is an unexplained paucity of F and G main sequence stars near the Sun with low Z values.

A Dissipative Collapse Model...
This involves the timescale of collapse, whether it is free-fall or dissipative...

The Hierarchical Merger Model
This is a more recent bottom-up approach that improves on the ability to explain age and metallicity
differences among the globular clusters. The proposal, fed by rising recognition of the importance of mergers,
is that ELS top-down formation takes place, but so does bottom-up via a hierarchy of mergers. In the latter,
disruption of globular clusters etc. leads to halo field stars... This would most strongly affect the center of
the Galaxy, building the bulge more rapidly than the halo. Perhaps 90% of original clusters have been
disrupted, especially the lower-mass ones and the high mass ones... The dSph galaxies may be proto-Galactic
remnants.

Formation of the Thick and Thin Disks
An attempt is made to deduce the gravitational potential energy as a function of z-displacement and directed
perpendicular to the disk, and thereby the local z-directed gravitational acceleration. This leads to the vertical
scale height of the thick disk.
Thin disk forms after the thick disk. A feedback loop operates: If the thin disk becomes too thin, the mass
density increases, supernovae increase, and the disk is reheated causing expansion. Most ongoing star
formation occurs in a young, inner portion of the thin disk with scale height of < 100 pc.

536
Young Stars in the Central Bulge

These arise from recent mergers with gas-rich satellite galaxiestidal interactions strip the gas, which settles
into the disk, forming new stars...

Metallicity Gradients...
Hierarchical mergers predicts metallicity gradients in galaxies that have undergone a dissipative collapse.
Such galaxies should exhibit a color gradient, namely that stars further from the center should be redder.
However, this patterns could be disrupted by mergers etc., and in fact starburst galaxies have inverted color
gradients and appear bluer in their centers.

Formation of Elliptical Galaxies
Many ellipticals have probably formed their stars early in the galaxy-formation process before a disk could
form. Other ellipticals have formed from mergers of spirals that would also disrupt disks and relax to the
elliptical pattern with an r
1/4
profile. Not fully explained is the distribution of globular clusters. Mechanisms
are being evaluated with N-body simulations. A competing hypothesis is that ellipticals develop in the
bottoms of the potential wells, even without interactions...

Galaxy Formation in the Early Universe
Early galaxy formation can be witnessed in distant galaxies. At 1 Mpc, the light we see was emitted 3 million
years ago.
According to the 1978 Butcher-Oemler effect
537
, galaxies used to be bluer than they are today (suggesting
more active new star formation back then). For more distant and therefore views of older times, there are
fewer elliptical and lenticular galaxies relative to spirals, consistent with their later formation by mergers.
The HST 2004 Ultra Deep Field image was of only 3 arcmin square in Fornax, and includes some galaxies
from only 400 to 800 Myr after the Big Bang. These distant galaxies exhibit an abundance of strange shapes
and blue galaxies that may be the building blocks for the more regularly shaped galaxies we see closer to
home.
The clumping of dark matter forming potential wells was likely critical to the formation of clusters of galaxies
(IMA2 p. 1030) and will be discussed in Chap. 30.

537
http://en.wikipedia.org/wiki/Butcher-Oemler_Effect
The Structure of the Universe (Chapter 27)

Summary of Distance Measuring Methods and The Extragalactic Distance Scale

The two dimensions of an objects location on the celestial sphere (or surface) are easy to measure, but the
astrometric measurement of the third dimension, distance, is much more difficult. There are a number of
methods that estimate distance to objects in astronomy, and they differ in applicable scale. The sequence of
methods that apply at differing but overlapping distances establishes a cosmological distance ladder which
is used to estimate distances to the farthest objects known.

Radar Ranging
On the closest scale, direct measurement of distance may be done precisely with radar ranging. This
technique however is confined to objects in the solar system, as the transmitted plus reflected signal exhibit a
combined fall-off 1/d
4
. Objects ranged have included Mercury, Venus, Moon, Mars, Jupiter and the
Galilean satellites, Saturn including its rings and Titan, and a number of comets and asteroids.
538
Saturn is
at about 9 AU or about 1.3 x 10
-4
LY, so that we may take 10
-4
LY as the maximal distance measurements of
this method.

Geometric or Trigonometric Parallax
This has been previously discussed (see Stellar Parallax). It is one of the gold standards outside the solar
system, especially as performed by the highly precise Hipparcos Space Astrometry Mission (High precision
parallax collecting satellite, 19891993).
539
Even better is expected from the upcoming and previously
mentioned GAIA mission, slated to launch in 2013.
540
Current maximal distance measurements with
geometric parallax by Hipparcos has been 1 milliarcsec, equivalent to 1000 pc or 3000 ly, but more commonly
ground measured distances to about 100 ly (including the Hyades star cluster
541
at 153 ly) are determined by
this method.
542
It is said that GAIA will measure down to 10 or 20 microarcsec, the latter presenting 50 kpc =
163,000 ly (IMA2 p. 59 and here
543
).

Main-Sequence Fitting (Spectroscopic parallax)
This has been previously mentioned. It applies primarily to clusters of stars, for which a valid sample of stars
are available at about the same distance. By comparing their H-R diagram to the H-R diagram of the absolute
magnitude of well-established main-sequence stars (for example the Hyades cluster, and correcting
presumably for any small amounts of redshift present), a displacement in magnitude may be observed
between the two curves. The difference between the line of apparent magnitude of the cluster and the line of
absolute magnitude of the main sequence line is the distance modulus used in the main sequence fitting.
(Problems can arise from blue stragglers, older clusters that have few main-sequence stars, and juxtaposed
stars that are not gravitationally bound in the cluster.) IMA2 p. 1038 says this method extends to about 7
kpc or 2,000 ly, and ZXI suggests this technique is useful to about 10
5
ly. The open star cluster M13

538
http://en.wikipedia.org/wiki/Radar_astronomy
539
http://en.wikipedia.org/wiki/Hipparcos
540
541
http://en.wikipedia.org/wiki/Hyades_%28star_cluster%29
542
http://outreach.atnf.csiro.au/education/senior/astrophysics/parallaxlimits.html
543
http://en.wikipedia.org/wiki/Gaia_%28spacecraft%29
(NGC 6205) is estimated by ZXI
544
to be at 32,600 ly by this technique, whereas a recently published value is
6.8 kpc = 22,200 ly.
545

Wilson-Bappu Effect
Another method involving analysis of stellar spectra is takes advantage of the Wilson-Bappu effect, in which
widening of spectral lines correlates with absolute magnitude. For example, the Ca II K absorption line shows
a well-defined centrally superimposed emission core with a quite narrow width that varies with absolute
magnitude. The technique is useful over a range of 15 mag (IMA2 p. 1039). It is calibrated with nearby stars.
Unfortunately, the dispersion of values is about 0.5 mag, rendering it too imprecise to significantly improve
the cosmic distance ladder.
546

Cepheid (and RR Lyrae) Variable Stars
Cepheids have been extensively discussed, in particular the Period-Luminosity-Color (PLC) relation. These
supergiant stars are one of the gold standards for galaxies and clusters containing these stars (the Cepheids
must be are individually identifiable). The technique is also called secular parallax (a misnomer). Cepheids
have been used to measure distance to M31, M100 (55 Mly), and other relatively nearby galaxies. Extinction
problems are best addressed by observing in infrared wavelengths.
The much smaller older and fainter metal poor Type II cepheids must be distinguished from the supergiant
metal-rich Classical Type I Cepheids.
The most distant known Cepheids (IMA2 p. 1041) are in NGC 3370
547
at 29 Mpc or about 10
8
ly. ZXI says the
method is useful to about 10
7
ly. Cepheids are an important method that spans the range of about 100 to
10
7
ly. The uncertainty is 0.4 mag (IMA2 p. 1051) or, from another source, about 7% (for nearby stars) to up
to 15% (for distant stars).
548

RR Lyrae variable stars also may be used to estimate distance because their luminosities are all about the
same. They are more common than Cepheids, are metal poor Population II horizontal branch stars found in
globular clusters, but also much less luminous than Cepheids..

544
http://www.astro.washington.edu/users/ivezic/REU08/GDP/distance/background4.html
545
Paust NEQ et al, The ACS Survey Of Galactic Globular Clusters. VIII. Effects of Environment on Globular
Cluster Global Mass Functions, ApJ Volume 139 Number 2 , 2010
546
http://en.wikipedia.org/wiki/Wilson%E2%80%93Bappu_effect
547
http://heritage.stsci.edu/2003/24/supplemental.html
548
http://en.wikipedia.org/wiki/Cosmic_distance_ladder#Classical_Cepheids
Supernovae Type Ia
Core-collapse supernovae (SNe Types Ib, Ic, and II)
have been used to estimate luminosity and
therefore distance, but the uncertainties are
highabout 25% for determining distance to the
Virgo cluster (IMA2 p. 1042)and Type II SNe at
least are dimmer than Type Ias.
However, the light curves of Type Ia SNe can be
much more tightly fit to a calibration curve. These
SNe are the best distance-determining standard
candles for distances to greater than 1000 Mpc (3
billion ly). These highly uniform SNe occur when a
binary white dwarf star, accreting matter from its
companion red dwarf star, reaches the
Chandrasekhar Limit of 1.4 M. Once this critical
mass is reached, the star undergoes a runaway
nuclear fusion reaction and explodes. Because all
Type Ia supernovae explode at about the same
mass, their absolute magnitudes are all the
same.
549
The absolute magnitude at peak of Type
Ia SN has been found to be
MB MV 19.3 0.3.
In the upper graph to the right,
550
the absolute
blue magnitude MB is plotted over 80 days (in the
stars rest frame). Most of the stars fall in the
narrow yellow band (not all stars are plotted), and
there are few outliers. The lower graph is the
same light curves rescaled. Perlmutter 2003 states The nesting of the light curves suggests that one can
deduce the intrinsic brightness of an outlier from its time scale. The brightest supernovae wax and wane more
slowly than the faintest... Simply by stretching the time scales of individual light curves to fit the norm, and
then scaling the brightness by an amount determined by the required time stretch, one gets all the type Ia
light curves to match.
Catching the SN at peak luminosity is desirable but not apparently essential for inferring the peak luminosity.
This inference involves either the aforementioned stretch method, or the multicolor light curve shape
method (MLCS), in which the shape of the light curve (taken after the initial explosion) is compared to a
family of parameterized template curves that will allow determination of the absolute magnitude at peak
brightness. The latter has the advantage of allowing for dust-related extinction and reddening of transmitted
light.
Once the scaled absolute blue magnitude and the apparent blue peak magnitude is determined, the distance
is calculated as usual with
d = 100
(m M A + 5)/5
pc
Type Ia SNe are about 13.3 magnitudes brighter than Cepheid variable, thus the technique is capable of
reaching out more than 500 times further. The uncertainty in distance modulus is about 0.1 mag, and the
uncertainty overall in distance estimation by this method is as low as 5%. The range is > 1000 Mpc. (IMA2 p.
1044)
However, Type IA supernovae (SNe) are quite rare. The last observed SN of any type in the Milky Way was the
1604 Keplers Supernova (SN 1604, type uncertain but apparently a Type IA).
551
Before Keplers SN, another
Type IA SN, SN 1572 (Tychos Supernova), occurred in 1572.
552

549
http://en.wikipedia.org/wiki/Cosmic_distance_ladder#Type_Ia_light_curves
550
Perlmutter, Saul, Supernovae, Dark Energy, and the Accelerating Universe, Physics Today, 56, April, 53
(2003).
551
http://chandra.harvard.edu/photo/2007/kepler/index.html
552
http://en.wikipedia.org/wiki/SN_1572
One volumetric survey estimate,
553
based on the SN Ia rate found from the Sloan Digital Sky Survey II (SDSS-
II) Supernova Survey (which found 516 SN Ias at redshift z < 0.3), is that the SN rate at a mean redshift of z =
0.12 is
rV = 2.69 10
5
SN yr
1
Mpc
3
(H0/(70 km s
1
Mpc
1
))
3

Novae
These are more common and may also be used like SNe to estimate distance (IMA2 p. 1044), typically using
the time from peak MV to 2 magnitudes less. This method extends to 20 Mpc, just past the distance to the
center of the Virgo cluster and similar to the maximal distance for Cepheids. The uncertainty is 0.4 mag.
Of course, the range is much smaller than Type Ia SNe.

Brightest Giant H II Regions; Brightest Red Supergiants
These can be used to provide a secondary standard candle (i.e., one requiring a galaxy at known distance for
calibration. Angular size vs. galaxy apparent magnitude are measured for the object and compared to the
same quantities in galaxies with known distance, etc. This method is not very accurate.
Similarly, the 3 brightest individually resolved Red Supergiants in a galaxy seem to have about the same
absolute MV = -8.0, and this can be the basis of distance estimate. This method is limited to about 7 Mpc.

553
Dilday B, et al, Measurements of the Rate of Type Ia Supernovae at Redshift z < ~0.3 from the SDSS-II
Supernova Survey, ApJ 713:1026-1036, 2010, http://arxiv.org/abs/1001.4995 Error range omitted.
Globular Cluster Luminosity Function GCLF
This statistical method is more accurate than measuring the top 3 red supergiants, and works best for giant
ellipticals with many globular clusters. For a globular star cluster, the uncertainty is about 0.4 mag, and the
method reaches out to 50 Mpc. However, uncertainties remain because there is no universal globular cluster
luminosity. The graph that follows
554
demonstrates the peak number of GCs found over all luminosities for
GCs around 4 elliptical galaxies in the Virgo Cluster, as well as the standard deviation in the peak.
The turnover absolute magnitude has been found with other samples to be M0 6.6.

This technique has an uncertainty is 0.4 mag (IMA2 p. 1051).

554
Jacoby GH, A Critical Review of Selected Techniques For Measuring Extragalactic Distances, Publications
of the Astronomical Society of the Pacific, 104: 599-662, 1992 August,
http://ned.ipac.caltech.edu/level5/Jacoby/frames.html
Planetary Nebula Luminosity Function PNLF
A similar statistical approach can be used with a collection of a galaxys planetary nebulae. The value
M5007 = 4.53 expresses the cutoff (maximum) absolute magnitude at the 500.7 nm [O III] forbidden line found
in all Population II PNs, a value which may be used as the standard candle. This is reliable in finding
distances to elliptical galaxies up to 20 Mpc (IMA2 p. 1046, see also here
555
) or up to 50 Mpc (IMA2 p. 1051).
The uncertainty is 0.3 mag (IMA2 p. 1051)

Surface Brightness Fluctuation Method
This technique looks at a galaxy as a whole, and can be used up to 100 Mpc, 125 Mpc with HST. It is best
done with a CCD detector. The amount of rms pixel to pixel variation in brightness is inversely proportional
to distance. (IMA2 p. 1047 and here
556
) This technique has an uncertainty is 0.3 mag (IMA2 p. 1051).

Tully-Fisher Relation (Spirals V
max
L)
Discussed previously, applies to spiral galaxies, relates spiral galaxy luminosity to maximum rotation velocity.
It is accurate (0.4 mag in IR). Despite its great range, > 100 Mpc, nearby spirals can also be calibrated with
Cepheids. This has helped to map the 3-D dimensions of the Virgo and Coma Clusters. It is widely used in
measuring extragalactic distances. (IMA2 p. 1048 and here
557
) This technique has an uncertainty is 0.4 mag
(IMA2 p. 1051).

D versus Relation (D )
This applies to elliptical galaxies. As previously mentioned, the Faber-Jackson relation
correlating
luminosity and velocity dispersion 0 is subject to excessive scatter for elliptical galaxies and not useful for
estimating distance. However, the relation between galaxy diameter D and velocity dispersion for a galaxys
stars is tighter. D here is the angular diameter out to a surface brightness (isophote) of 20.75 B-mag arcsec
-2
.
D is inversely proportional to the galaxys distance. This method then give a standard ruler rather than a
standard candle, in that a measured 0 predicts a certain diameter D, by log10D = 1.333 log10 + C. Here, C
depends of the distance to the galaxy cluster. This is also useful for determining relative distances between
two galaxy clusters. It can be used to show that Coma Cluster is more than 5 times as far as the Virgo
Cluster. However, this method has not yet been calibrated with Cepheids. (IMA2 p. 1048 and here
558
) This
technique has an uncertainty is 0.5 mag (IMA2 p. 1051) and a range of > 100 Mpc.
The term fundamental plane is sometimes used for this technique as applied to normal elliptical galaxies.
The plane in question is not a normal spatial plane, but a 2-dimensional abstract plane within a 3-
dimensional abstract space having coordinates of effective radius, average surface brightness, and central
velocity dispersion. Any one of the three parameters may be estimated from the other two, as together they
describe a plane that falls within their more general three-dimensional space.
559
The Luminosity may be
given by an empirical formula expressing it as a function of dispersion and effective radius. Alternatively, the
effective radius may be given by an empirical formula expressing it as a function of dispersion and surface
brightness (IMA2 p. 988)

555
Planetary nebula luminosity function:
http://en.wikipedia.org/wiki/Planetary_nebula_luminosity_function
see also http://arxiv.org/abs/0708.4292
556
http://ned.ipac.caltech.edu/level5/Jacoby/Jacoby9_1.html
557
Tully-Fisher Relation:
http://www.noao.edu/staff/shoko/tf.html
http://en.wikipedia.org/wiki/Tully%E2%80%93Fisher_relation
558
http://en.wikipedia.org/wiki/Cosmic_distance_ladder#D-.CF.83_relation
559
http://en.wikipedia.org/wiki/Fundamental_plane_%28elliptical_galaxies%29
Brightest Galaxies in Clusters (BCG)
The brightest galaxy in a cluster of galaxies can be used to obtain the clusters distance. The brightest value
has been found to average absolute visual magnitude MV = 22.83 0.61 mag, approximately 3 magnitudes
brighter than the peak brightness of Type Ia SNe. This method should be good to over 4000 Mpc, and thus
allow viewing of light emitted >13 billion years ago. However, galaxies may have evolved in intrinsic
brightness over cosmological time periods, and galaxies at great distances may be different, for instance by
mergers, or they may be smaller,
560
so caution is needed in applying this method.

The Expansion of the Universe

Vesto M Slipher (18751969) first discovered in 1914 that most nebulae were redshifted and therefore moving
away from us. (although the relatively nearby Andromeda M31 galaxy is blue shifted). He also found that
most nebulae are moving apart from each other, indicating a general expansion of the universe.

Hubble Law
Hubble discovered Cepheids in M31 in 1925. His publication in 1929 included the Hubble Law or Relation:
the velocity of distant galaxy recession is proportional to the distance. This relation is now expressed in non-
relativistic form as v=H0d. The distance d is given in Mpc and recession velocity v in km s
-1
, so H0 has units
of km s
-1
Mpc
-1
. He and assistant Milton Humason compiled many distances and redshifts, working for
instance with the H and K lines of Ca II.
The factor H0 applies only to the present time and is commonly expressed as 100h km s
-1
Mpc
-1
, where h is a
dimensionless multiplier derived from various sources. (It is of course not the same as the Planck constant
h). The WMAP value of h, namely hWMAP, is currently accepted as one the best estimates:

Therefore, when z << 1, d cz/100h Mpc, or d cz/71 Mpc. This is the nave Hubble distance,
applicable only for small z.

Recessional Velocity
According to IMA2 p. 1054, a galaxys recessional velocity is not due to its movement through space but
instead to an expansion of space itself which carries the galaxy along. Thus it differs from peculiar velocity.
This motion of the galaxies due to expansion is called the Hubble Flow. Note that gravitationally bound
systems do not expand with the Hubble Flow (IMA2 p. 1056).

Cosmological Redshift
The redshift attributable to expansion of the universe is called the cosmological redshift, and is due to the
corresponding expansion of the wavelength of the light emittedthus it is not accurate per General Relativity
to use the Doppler shift equations developed for a static Euclidean universe. However, astronomer s use the
relativistic Doppler formula for convenience to translate a measured redshift parameter z into a peculiar
velocity that would apply if the redshift were due to relative motion rather than to the expansion of spacetime.
Cosmological redshift attributable to Hubble expansion is so named to distinguish it from more familiar
Doppler redshift (due to peculiar velocities of galaxies, etc.) and from gravitational redshift (arising when light
climbs out of a deep potential well).
The following is an abbreviated discussion.
Recall that redshift parameter z is defined:

560
http://www.nasa.gov/mission_pages/hubble/science/far-protocluster.html

z = (obsv - rest) / rest = / rest = obsv / rest - 1 = vr/c

or

z = (frest fobsv ) / fobsv = f / fobsv = frest / fobsv - 1 = vr/c
The relativistic Doppler shift is given by
/c
+v
/c

where vr is the observed radial velocity, vr = u cos , and = angle between line to observer and the direction
of movement at velocity u.
The redshift for purely radial motion is given as

+v
/c
v
/c

and distance is given approximately by

c
+

When z << 1, d = cz/H0.
An excellent generalrelativityaware JavaScript web-page calculator by EL Wright is found here.
561
One
inserts z (plus optionally values for H0, M, vac, and whether ) and computes comoving radial distance, time
since the Big Bang, and light travel (elapsed) time.
The following graphs show various distances in Gly (or age in Ga for lookback time) as a function of
cosmological redshift z:
562
(Graph curves in the left diagram are named from top to bottom in the legend at
lower right of this graph.) Note that the Nave Hubble graph (dot-dash line) is always a straight line, because
for this graph Dnaive=cz/H0. Note log-log scales for second graph axes. Here,
Nave Hubble Dnaive=cz/H0.
LOS comoving D= line-of-sight comoving radial distance=proper motion distance;
Luminosity DL based on luminosity vs. flux for known object, where Flux = Luminosity/(4*pi*DL2)
where DL = DA*(1+z)2
Look back time tL=light travel distance = c(t0-temit) (hard to determine because temit uncertain).
Angular diameter DA= angular size distance = x/ (x=actual size, = angular size), DL = DA*(1+z)2

561
http://www.astro.ucla.edu/~wright/CosmoCalc.html, which is described in
Wright Edward L., A Cosmology Calculator for the World Wide Web, The Publications of the Astronomical
Society of the Pacific, Volume 118, Issue 850, pp. 1711-1715, 12/2006. The article spells out how the general
relativistic calculations are made. I have not studied general relativity and have not taken the time to review
the details of these calculations.
562
http://en.wikipedia.org/wiki/Distance_measures_%28cosmology%29
See webpage for detailed explanation, not all of which I understand. See also and
http://www.astro.ucla.edu/~wright/cosmo_02.htm and http://www.atlasoftheuniverse.com/redshift.html

Another graph follows, showing the full range of cosmological redshifts for objects currently visible with HST
(about z = 15, neglecting the CMB at z 1100):

Graph from Richard Collins
563
modified by MCM
Note linear scales for axes.
DL = Luminosity Distance, DC = Comoving Distance,
DT = Light Travel Time Distance (Lookback Time distance), and DA = Angular Diameter Distance

563
http://www.atlasoftheuniverse.com/redshift.html
The Comoving Distance [DC] is the distance scale that expands with the universe. It tells us where the
galaxies are now even though our view of the distant universe is when it was much younger and smaller [and
the object much closer]. On this scale the very edge of the visible universe is now about 47 billion light years
from us although the most distant galaxies visible in the Hubble Space Telescope will now be about 32 billion
light years from us.
564

Big Bang and Cosmic Microwave Background CMB
The Big Bang is the inferred past beginning of an explosion taking place where all matter and space was
consolidated into a tiny size, perhaps a mere point. The Big Bang left a remnant in the form of blackbody
radiation of this intensely hot explosion. Due to the expansion of space as the light from this event has
propagated, the wavelengths of this intensely energetic light emitted have become elongated, causing an
effective cooling of the radiation, so that it now has a blackbody teff = 2.725 K (graph from NASA
565
follows).

If we assume that the expansion has proceeded at uniform velocity (actually an invalid assumption) we get a
crude estimate of the Hubble Age or Hubble Time tH, the time it takes to play the Big Bang explosion in
reverse back to its origin. This time tH = 13.8 Gyr. (IMA2 p. 1058)

Groups and Clusters of Galaxies

The universe is still not of uniform structure at a scale somewhat greater than that of galaxies. Instead,
galaxies often occur gravitationally bound and orbiting the barycenter of galaxy groups and clusters. The
structure of the universe consists of galaxies, groups of galaxies, clusters of galaxies, and clusters of clusters
(Superclusters).
Groups of Galaxies: are the smallest bound aggregate of galaxies, with typically no more than 50 galaxies in
a diameter of 1.4h
-1
Mpc and an average mass of approximately 2x10
13
h
-1
M. The spread of velocities for the

564
ibid.
565
http://map.gsfc.nasa.gov/universe/bb_tests_cmb.html
individual galaxies is about 150 km/s. Larger and more massive galaxy systems are sometimes also classified
as galaxy groups. The mass to light ratio is typically about 260h M/L, indicating much dark matter.

The Milky Way is contained in the Local
Group of more than 54 galaxies, see below.
(IMA2 p. 1058 and here
566
)

The graph to the right depicts the relative 3D
positions of the galaxies in the Local Group.
567

The graph is centered on the MW, but the MW
is not at the barycenter (it is 462 kpc from the
MW). Note the wide separation between MW
and M31 (Andromeda Galaxy).

Other groups are perhaps less well known, but
include the IC 342/Maffei Group, the M81
Group, and the Centaurus A/M83 Group.
(The two groups with / in the name each
consist of 2 subgroups, thus the complex
naming). The Sculptor group is the closest
group at 1.8 Mpc, and the M81 group is at 3.1
Mpc. Several additional groups are within 10 Mpc of the Local Group.

Galaxy Clusters: are larger than groups, although there is no sharp dividing line between the two. Clusters
hold from 50 (a poor cluster) to up to many thousands of galaxies (a rich cluster) within a span of about 6h
-1

Mpc, and with a typical mass of 1x10
15
h
-1
M. Velocity dispersions are 800 to up to 1000 km s
-1
. The mass
to light ratio is typically about 400h M/L, again implying much dark matter. Clusters can be spherical or
irregular. (IMA2 p. 1058). Most galaxies are found in small groups or poor cluster (akin to the fact that there
are more dwarfs than giant galaxies). (IMA2 p. 1061)
Notable nearby galaxy clusters include the Virgo cluster, Fornax Cluster, Hercules Cluster, and the Coma
Cluster.
568

Zero Velocity Surface
The edge of a cluster or group is not well defined visually or otherwise, and thus the concept of zero velocity
surface ZVS is useful to define their spatial limits. (The term ZVS arises in the problem of N-body orbits.) It
is especially useful with loosely bound orbiting bodies such as galaxies. The ZVS is the mathematically-
defined 3-D surface that is the theoretical eventual limit of excursions of the galaxies in the cluster, given
their current velocities and directions. In other words, it defines a surface boundary that the galaxies cannot
cross. (IMA2 p. 1059 and here
569
, a complex topic that I have not studied)

The Local Group
This is more than 54 nearby galaxies, located within about 1 Mpc from the MW. The ZVS is at about 1.2 Mpc
from the barycenter. The Local Group includes the MW, the LMC, the SMC, M31 (Andromeda), the
Triangulum Galaxy, and an assortment of dwarf ellipticals and dwarf spheroidals, etc., some of which are in
the Magellanic Stream. It is part of the Virgo Supercluster. The MW and M31 are the largest galaxies and

566
http://en.wikipedia.org/wiki/Galaxy_groups_and_clusters
567
http://en.wikipedia.org/wiki/Local_Group
568
http://en.wikipedia.org/wiki/Galaxy_cluster
569
http://farside.ph.utexas.edu/teaching/336k/newton/node125.html
widely separated, altogether providing 90% of the luminosity of the group. The Local Group is in a state of
collapse, and in particular the MW and M31 will collide (but only after 6.3 Ga), perhaps forming a giant
elliptical. It is inferred that after the Big Bang, the galaxies of the Local Group were initially in recession from
each other, but gravitational attraction eventually stopped this expansion, and they began approaching each
other. The total mass of MW and M31 combined is calculated by orbital considerations to be 4x10
12
M.
Their mass to light ratio is about 57 M/L but luminous matter (in the MW at least) only explains about 3
M/L, again implying much dark matter. (IMA2 p. 1060-1)

Virgo Cluster
This rich irregular cluster is in the constellations of Virgo and Coma Berenices. It covers 10 degrees square in
the heavens. The center is about 16 Mpc from us. It comprises about 2250 small and large galaxies in a
span of about 34 Mpc.
570
The four brightest galaxies are all giant elliptical galaxies, 3 of which are close to
the center of the cluster. These include M87 (Virgo A or NGC 4486), as well as M84 (NGC 4374) and M86
(NGC 4406).
M87 is a giant E1 elliptical active galaxy having a mass of about 6x10
12
M enclosed in 50 kpc radius, about
200 times that of the MW (IMA2 p. 1062 and here
571
). For M87, the enclosed mass Mr increases linearly with
radius out to about 300 kpc, a finding consistent with dark matter accounting for 99% of the mass in this
galaxy. (IMA2 p. 1063)

Coma Cluster (Abell 1656)
This is the nearest rich and regular cluster, located in the constellation Coma Berenices. It has about 10,000
galaxies (most of which are dwarf ellipticals and S0s, with only 15% spirals and irregulars). It is about 90
Mpc away, about 5 times further than the Virgo Cluster. It is at a redshift z=0.0231, yielding a light travel
time of about 0.3 Ga, and has an estimated mass = 7x10
14
M. Its brightest members are the giant cD
ellipticals NGC 4874 and NGC 4889, both near the center. Younger spirals are few and probably
peripheral.
572
Fritz Zwicky first observed that there is not nearly enough luminous matter to account for
observed orbits of the Coma galaxy cluster, and posited dark matter. The mass to light ratio can be
simplistically calculated at about 660 M/L (but actually is about half that, IMA2 p. 1066)

Evolution of Galaxies
By analysis of distant and therefore older galaxy clusters , such as the rich cluster CL 0939+4713 (Abell 851)
at z=0.41 (distance = 6.4h
-1
corresponding to a view from 9 Ga ago), we can infer how galaxies in clusters may
have evolved. For this much older cluster, many spirals are seen in the center, whereas in the much more
recent views of Coma Cluster, few central spirals are seen. (IMA2 p. 1064) It is theorized that the ellipticals
form from tidal interactions and hierarchical protogalaxy mergers, leaving ellipticals and SOs.
An even more distant cluster , which includes 3C 324 and is in constellation Serpens, has z=1.2 and is
therefore viewed at 9 Ga ago.
573
It shows a much more disorderly arrangement of galactic fragments, perhaps
representing future or former spirals. The HubbleSite description
574
states: Very few of the cluster's
members are recognizable as normal spiral galaxies (like our Milky Way), although some elongated members
might be edge-on disks. Among this zoo of odd galaxies are tadpole-like objects, disturbed and apparently
merging systems dubbed train-wrecks, and a multitude of faint, tiny shards and fragments, dwarf galaxies or
possibly an unknown population of objects. However, the cluster also contains red galaxies that resemble
mature examples of today's elliptical galaxies. Their red color comes from older stars that must have formed
shortly after the Big Bang.

570
http://en.wikipedia.org/wiki/Virgo_cluster
571
http://en.wikipedia.org/wiki/Messier_87
572
http://en.wikipedia.org/wiki/Coma_Cluster
573
574

Hot Intergalactic Intracluster Gas and Intercluster Medium ICM
Some (but not nearly all) of the missing mass per Zwickys calculation of mass to luminosity relationship is
attributable to intracluster gas in the intracluster medium ICM. This is very hot, fully ionized, and emits in
the X-ray wavelengths, and filling the cluster gravitational potential wells. The X-ray luminosities for cluster
ICM are 10
36
to 10
38
W. The Virgo and Coma Clusters have, resp., about 5 x10
13
and 3x10
13
M in ICM gas
mass. The radiation is produced by the thermal bremsstrahlung process (free-free radiation), where the
bremsstrahlung X-rays are created by the acceleration of free electrons in the electric fields of ions (mostly
protons) in a plasma. The reaction is diagrammed as:
e + p e'+ p' +
According to ZXI, there is little relatively optical emission because there are no optical transitions and the gas
is optically thin (thus not like a black body).
For fully ionized H, the energy emitted by thermal bremsstrahlung per unit volume per unit time between
frequencies and + d is given by (IMA2 p. 1066):
T
/
e
/

and for all wavelengths is

An ESA/XMM image of the X-ray emission in the 0.3-2.0 keV energy band from the ICM of the Coma Cluster
follows.
575
The irregular distribution of the ICM may indicated mergers of several subclusters. The brightest

575
from the ESA/XMM-Newton mission, EPIC-pn camera, U. Briel, MPE Garching, Germany, color calibration
not provided, http://xmm.esac.esa.int/external/xmm_science/gallery/public/level3.php?id=47,
published in
region, shown in red and yellow, is centered on NGC 4874 and NGC 4889, giant cD elliptical galaxies, but
clearly there is much emission outside the galaxies that are indicated.

For the Coma Cluster, as observed with NASAs HEAO 1 A-2 (High Energy Astronomy Observatories, Cosmic
X-ray Experiment CXE) and OSO 8 (8th Orbiting Solar Observatory), the best fit for the observations as
thermal bremsstrahlung process yields effective temperature T = 7.6 keV = 88 x 10
6
K (here
576
and IMA2 p.
1068). The number of free electrons per m
3
is about 300, thus much less dense than giant molecular clouds
(as previously described). The total mass in the ICM, estimated by the formulas just given and assuming 1
proton per electron, is 1x10
14
M , which overestimates compared to the preferred estimate of 3x10
13
M.
Even the former is only a few percent of the total mass of the cluster.
The ICM also exhibits emission lines of highly ionized Fe, Si, and Ne, indicating that it has been enriched by
stellar nucleosynthesis, perhaps through dispersion from high dynamical friction during mergers, or by
bursts of star formation.

Briel UG, et al, A Mosaic of the Coma Cluster of Galaxies with XMM-Newton, Astronomy and Astrophysics,
v.365, p.L60-L66 (2001)
576
Henriksen MJ and Mushotzky RF, The X-ray spectrum of the Coma Cluster of galaxies, Astrophys. J. 302,
287, 1986
The Existence of Superclusters, Bubbles, Filaments, and Voids
Superclusters exist and are up to 100 Mpc in size. Our Local Supercluster contains the Virgo Cluster and,
near its edge, the Local Group. The latter is pulled toward constellation Virgo at a Virgocentric peculiar
velocity of 168 km s
-1
.
This is a minor perturbation of the much larger perturbation or inhomogeneity in the Hubble Flow from
expansion. This perturbation (gravity anomaly) appears to arise from a Great Attractor toward
constellation Centaurus (and located within the range of the Centaurus Supercluster
577
), much of which
appears to be dark matter. However, the existence of this attractor is controversial, though it seems likely
that there are perturbations in the Hubble Flow. (IMA2 p. 1070-1)
The Shapley Supercluster or Shapley Concentration
of galaxies is another postulated contributor of some of
the perturbation attributed to the Great Attractor. It
has a mass of ~10
16
h
-1
M and a distance of about 200
Mpc, so lies much further from the suspected location of
the Great Attractor. (see image to right)
Other nearby superclusters include the Hydra-
Centaurus Supercluster, the Perseus-Pisces
Supercluster, and the Coma Supercluster.
The CfA Homunculus lies at the heart of the Great
Wall, and the Coma Supercluster forms most of the
homunculus structure
578
(see below).
Between the superclusters, which in large scale plots
appear like clumpy filaments, are voids measuring up
to 100 Mpc. Galaxies have not had time to form in
these voids and then move out. The filaments appear
connected, like the areas of intersection in a froth of
soap bubbles.
The largest scale aggregations of clusters, seen in the 2dFGRS (2dF Galaxy Redshift Survey),
579
Sloan Digital
Sky Survey , and the Galaxy And Mass Assembly GAMA survey, are the Great Wall and the Southern Wall,
the largest known discrete structures in the universe. (IMA2 p. 10741080, further details omitted.)
At the very largest scales, the cosmological principle (a sacred cow assumption) dictates that there should be
no localized attractor that would impart a preferred direction in space.
The following nearInfrared panoramic views of galaxy clusters out to z=0.1 (comoving radial distance of 400
Mpc) of the entire near-infrared sky reveals the distribution of galaxies beyond the Milky Way. The image is
derived from the 2MASS (Two Micron All Sky Survey) Extended Source Catalog (XSC)more than 1.5 million
galaxies, and the Point Source Catalog (PSC [also from 2MASS])nearly 0.5 billion Milky Way stars. The
galaxies are color coded by redshift obtained from the UGC, CfA, Tully NBGC, LCRS, 2dF, 6dFGS, and SDSS
surveys (and from various observations compiled by the NASA Extragalactic Database), or photometrically
deduced from the K band (2.2 um [near-infrared]). Blue are the nearest sources (z < 0.01); green are at
moderate distances (0.01 < z < 0.04) and red are the most distant sources that 2MASS resolves (0.04 < z <
0.1). The map is projected with an equal area Aitoff [projection] in the Galactic system (Milky Way at
center).
580
I have included versions without and with labels for the clusters. A more accurate description of
the color coding is included just below, also from the Jarrett article. The article cited is also the source of the
Shapley Concentration image, above.

577
http://en.wikipedia.org/wiki/Great_Attractor
578
http://en.wikipedia.org/wiki/Galaxy_filament
579
http://www.mso.anu.edu.au/2dFGRS/
580
http://spider.ipac.caltech.edu/staff/jarrett/papers/LSS/ , published as
Jarrett TH, "Large Scale Structure in the Local Universe: The 2MASS Galaxy Catalog", 2004, PASA, 21, 396
Labeled image from http://en.wikipedia.org/wiki/Great_Attractor

Another galaxy survey, the 2dF Galaxy Redshift Survey (2dFGRS), provides us with the following
distribution of clusters out to z=0.25:
581

The galaxies cover an area of approximately 1500 square degrees at z<0.25
selected from the extended APM Galaxy Survey in three regions:
an NGP [part-way toward the North Galactic Pole] strip,
an SGP [part way to the South Galactic Pole] strip,
and random fields scattered around the SGP strip.

581
http://magnum.anu.edu.au/~TDFgg/ (image cropped by MCM)
Active Galaxies (Chapter 28)

This is a fascinating area of cutting-edge astronomy, and deserves far more extensive coverage than I have
been able to give to it here.

Classification and Characteristics of Active Galaxies

An active galaxy is a galaxy hosting an unusually active galactic nucleus (AGN). Active galaxies of a wide
variety of types have been described and the terminology has been confusing. An AGN may be loosely defined
as the central region of a galaxy which exhibits central luminosity (energy release) much higher than typical
for stellar processes. Another definition, which ignores the degree of luminosity, might be, A galaxy nucleus
which shows evidence for accretion onto a supermassive black hole [SMBH].
582
The presence of a SMBH is
not alone sufficientit must be actively accreting and emitting copious radiation of some form.
Active Galaxies are thought to have SMBHs in their nuclei. ZXI states (lecture 5) that most ellipticals and
bulges contain SMBHs.
In addition, the following features may be found. (This body of knowledge has been confusing but is clarified
by a unification theory, see later section):
Name or
Class
Radio
Q=Quiet
L=Loud
Characteristics
583

Seyfert 1
RQ
Seyfert Galaxy 1: Moderate overall luminosity; strong/visible blue optical
continuum; host galaxy clearly visible;
broad + narrow lines; narrow lines have high ionization; weak radio emissions;
moderately variable X-ray emissions; most are spiral galaxies; variable
Seyfert 2
RQ
Seyfert Galaxy 2: Similar to Seyfert 1; narrow emission lines but no broad lines;
weaker or no blue optical continuum; weak radio; weak X-ray emissions (possibly
due to absorption); most are spiral galaxies; not variable
QSR
RL
Radio-loud Quasar: Strong radio; strong optical continuum; broad + narrow
emission lines; some polarization; FRII; variable
QSO
RQ
Radio-quiet Quasar (Quasi Stellar Object): Broad + narrow emission lines; host
galaxy barely or not visible; strong optical continuum; weak polarization; variable
QSO-2
RQ
QSO-2: same as QSO but missing broad lines; not many currently known (some are
IRAS-QSOs)
LINER
RQ/L
Low Ionization Nuclear Emission-line Region galaxy: Weak/no continuum;
narrow lines of low ionization & moderate strength; can be either radio quiet or
loud;
BLRG
RL
Broad Line Radio Galaxy: Similar to Seyfert 1 but strong radio;
broad + narrow lines; strong radio emissions; FR II; weak polarization
elliptical galaxies; variable

582
http://www.astro.virginia.edu/class/whittle/astr553/Topic15/Lecture_15.html
583
Table adapted and revised by MCM from ibid. plus IMA2 p. 1107 etc.
NLRG
RL
Narrow Line Radio Galaxy: Similar to Seyfert 2 but strong radio;
narrow emission lines only; strong radio emissions; FR1 and FRII, no polarization
elliptical galaxies; not variable
PRG-II
RL
Powerful Radio Galaxy of Fanaroff-Riley class II (FRII): Edge brightened, powerful
jet; unspecified optical spectrum (could be BLRG/NLRG/LINER)
RG-I
RL
Radio Galaxy of Fanaroff-Riley class I (FRI): Similar to PRG-II except lower radio
luminosity & Fanaroff-Riley class I FRI (edge darkened, lower power jet)
LD-QSR
RL
Lobe Dominated Radio Loud Quasar: FR-II radio morphology; optically similar to
QSO
CD-QSR
RL Core Dominated Radio Loud Quasar: Flat spectrum; optically similar to QSO
BL-Lac
RL
BL Lacertae object: A type of Blazar. Strong featureless continuum, with no or
weak emission lines, highly and rapidly variable; strong polarization; strong radio
flat spectrum core; 90% are in ellipticals
OVV
RL
Optically Violently Variable Quasar: A type of Blazar. Similar to BL Lac but
normal QSO spectrum and much more luminous than BL-Lacs
Broad + narrow emission lines; strong radio; strong polarization; rapid variability.

Seyfert Galaxies
Seyfert galaxies (Types 1 and 2) are a class of galaxies with active galactic nuclei that produce spectral line
emissions from highly ionized gas, first observed in 1908 by Edward Fath (1880-1959) and named after Carl
Keenan Seyfert (1911-1960), who first described the class in 1943. The centers of Seyfert galaxies are
thought to contain SMBHs with masses between 10
7
and 10
8
M. Seyfert galaxies are characterized by
extremely bright nuclei, and spectra which have very bright emission lines of hydrogen, helium, nitrogen, and
oxygen. These emission lines exhibit strong Doppler broadening, which implies velocities from 500 to 4000
km/s, and are believed to originate near an accretion disk surrounding the central black hole.
584
The lines
are from both allowed and forbidden transitions. Seyfert II have only narrow emission lines. It is believed
that Type 1 and Type 2 galaxies are in essence the same, and they only differ due to the angle at which they
are observed. This is known as Seyfert Unification theory. In Type 2 Seyferts it is believed that the broad
component is obscured by dust and/or by our viewing angle on the galaxy.
585
Seyferts intermediate between
1 and 2 are classified as 1.5, etc. Seyferts show a featureless continuum devoid of lines [i.e., between what
lines are present], contributing much of the luminosity and often overwhelming the light of the galaxys stars
(IMA2 p. 1086) See also table.
Examples include Mrk 1242 and Mrk 1157 (as listed in E. Benjamin Markarians catalog) and NGC 4151 (the
Eye of Sauron, a nearby galaxy with readily discernible disk and bright nucleus).

584
http://en.wikipedia.org/wiki/Seyfert_galaxy
585
ibid.
Spectral Energy Distributions of Active Galactic Nuclei
The spectral energy distributions (SEDs, typically plotting flux density against frequency) of many types of
AGNs including quasars exhibit varying degrees of the following (summary mostly from IMA2 p. 1088).
Wide Range: A very broad energy distribution spanning 8 to 10 or 15 orders of magnitude of photon
frequency. (This is quite different from a pure Planck blackbody distribution of a star, which rises and
falls over a much narrower range.) AGN spectra are thought to be a combination of thermal
(blackbody, low polarization) and non-thermal emissions (synchrotron radiation, often polarized).
Radio: Either relatively radio-loud or radio-quiet
IR: A turnover (flattening off of the rise of the SED) at the infrared IR Bump The bump is a
thermal feature, thought to be emission from warm dust grains with T 2000K.
Light and UV: A Big Blue Bump from high UV output (a thermal component contributing much of
the bolometric luminosity of the source, perhaps from an optically thick accretion disk or free-free
(Bremsstrahlung) emissions).
X-ray: A wide ongoing continuum into X-ray and even Gamma-ray
586
photon energies
The SEDs of AGNs therefore do not follow a simple power law, which would be expressed by

where F is flux density for frequency , and = spectral index, but such a relationship may be useful to
parameterize the continuum. Such power law behavior would be characteristic of synchrotron radiation
generated by relativistic electrons circling magnetic field lines under the Lorentz force.
Here are representative SEDs for quasars of radio loud and radio quiet type (image annotated by MCM):
587

586
http://imagine.gsfc.nasa.gov/docs/science/know_l1/active_galaxies.html
587
Zhaohui Shang, et al, The New Generation Atlas of Quasar Spectral Energy Distributions from Radio to X-
rays, http://arxiv.org/abs/1107.1855 (image annotated by MCM)
Such SEDs over a large range of frequencies draw on data from many sources: VLA, Submillimeter array,
FIRST, Spitzer, IRAS, ISO, 2MASS, SDSS, HST, FUSE, GALEX, ROSAT, Chandra, XMM, and Compton, etc.
(ZXI lecture 7 and here
588
)

Radio Galaxies
The first found, Cygnus A (3C 405)
589
, was discovered by Grote Reber (1911-2002) in 1939. Karl Jansky
(1905-1950) developed the early field of radio astronomy. Cyg A is the brightest extragalactic radio source
and only exceeded in output by the Sun and Cassiopeia A in the MW (Sagittarius A* at the center of the MW is
not strong). Baade and Minkowski (1895-1976) were able to locate an optical counterpart for Cyg Ait is a
cD galaxy whose nucleus is surrounded by dust. Its distance is ~240 Mpc, so that the radio emission is
extremely bright, the radio luminosity estimated by IMA2 p. 1092 at Lradio = 4.8x10
37
W. (See image below.)
Radio Galaxies have also been divided into Broad Line Radio Galaxy (BLRGs, similar to Seyfert 1s but radio
loud) and Narrow Line Radio Galaxy (NLRG , similar to Seyfert 2 but radio-loud, giant or supergiant
ellipticals, include Cyg A).
One of the largest radio galaxies is 3C 236, spanning about 4.5 Mpcit is a FR II radio galaxy.
590

A nearby example is Centaurus A (NGC 5128), which produces a prominent radio- and X-ray-emitting jet.
The galaxy is hard to categorizeit is listed by NED as a peculiar S0 lenticular galaxy and by Simbad as a
Seyfert 2 galaxy. (See image below.)

Radio Lobes and Jets
Radio galaxies may emit radio from extended lobes fed by jets or from a compact core or a halo. The jets tend
to be strongly collimated (see model discussion below). At least half of the stronger radio sources have
observable jets.
The most powerful radio galaxies often appear to have one-sided jets, but this is often an effect of relativistic
beaming (headlight phenomenon) or extinction of the counterjet.
High resolution imaging of the jet regions, e.g., of NGC 6251, shows that the jet can be traced back to the core
of the galaxy. Jets often display knots or hot spots of enhanced luminosity (in radio, visible, or X-ray)
suggesting variable outputs of material. M87 (Messier 87, aka Virgo A) is one of the brightest radio objects in
the sky, and exhibits a highly collimated jet with knots, extending out 1.5 kpc from the core of its supergiant
elliptical galaxy.
591
The X-ray luminosity of M87 overall is 50 times the radio luminosity.
Jets do not always remain straight, and may appear curved or windblown, apparently distorted by the
galaxys movement causing friction with its intracluster gas. See for example, NGC 1265 which is moving
through the Perseus cluster (Abell 426).
592

588
Zhaohui Shang, 2009, http://202.127.29.4/cosmology09/talk/ShangZhaohui.ppt
589
The 3C catalog is Third Cambridge Catalogue of Radio Sources, http://en.wikipedia.org/wiki/3C_Catalog
590
Tremblay GR; et al; Episodic Star Formation Coupled to Reignition of Radio Activity in 3C 236, ApJ
Volume 715, Issue 1, pp. 172-185 (2010)
591
http://en.wikipedia.org/wiki/Messier_87
592
Abell = Abell catalog of rich clusters of galaxies, http://en.wikipedia.org/wiki/Abell_catalogue

Centaurus A (NGC 5128), superimposed dual radio lobes (VLA) and optical (probably HST),
Dark dust lanes partially obscure the optical elliptical galaxy.
Distance is 10 Mly, image spans 18.2 arcmin, 1998
593

Cygnus A, showing central active galaxy (bright dot), collimated highly asymmetrical jets, and radio lobes.
5 GHz, 0.5'' resolution, 0.03 FOV, redshift z = 0.057, d= 230 Mpc. 1983 (image: NRAO/AUI)
594

593
http://hubblesite.org/newscenter/archive/releases/1998/14/image/l/
594
http://images.nrao.edu/110
Quasars
Much of what follows under this section applies to galaxies with AGNs, whether or not they are classical
radio-loud quasars.
A quasi-stellar radio source (quasar) is a very energetic and distant active galactic nucleus. Quasars are
extremely luminous and often highly redshifted sources of EM energy, including radio waves and visible light.
They appear point-like, similar to stars, rather than as more extended galaxy-like sources or radio lobes.
Quasars are all believed to have a SMBH at the center. The size of a quasar is 1010,000 times the
Schwarzschild radius of the black hole. The quasars emissions arise from the accretion disc.
The term quasar has undergone evolution and can be ambiguous. It originally described radio-loud sources
(now usually abbreviated QSRs), but quasar has been extended to include radio-quiet quasars (quasi-stellar
objects, now abbreviated QSOs). The name and abbreviation quasi-stellar objects or QSO are also used to
indicate quasars of either type.
The first observed in optical was by Allan Sandage and Thomas Matthews in 1960. They were searching for a
counterpart to the radio source 3C 48. This faint quasi-stellar object was emitting more intense radio waves
and ultraviolet radiation than a typical star. Its SED revealed a unique spectrum displaying broad emission
lines that could not be identified with any known element. A similar radio-loud quasar was found for 3C 273,
which happened to have a prominent jet and a redshift z = 0.16, making spectroscopic identification of lines
difficult. The absolute visual magnitude MV was calculated as 26.2, yielding a visual luminosity of 10
39
W.
The radio luminosity was about 7 x 10
36
W and the bolometric luminosity is estimated at 2.2 x 10
40
W.
595
The
observed range of bolometric luminosities for quasars is 10
38
to 10
41
W (IMA2 p. 1097). The most energetic
quasars have 10
5
times the bolometric luminosity of normal galaxies like the Milky Way.
The redshift of 3C 48 was subsequently found to be z = 0.367, corresponding to a radial speed of 0.303c and
a Hubble distance of 900h
-1
Mpc, making it one of the most distant of known objects.
The most distant quasar yet discovered by the UKIDSS Large Area Survey ULAS
596
is ULAS J1120+0641 with
a redshift z = 7.085, thus seen only some 770 million years after the Big Bang. It took 12.9 billion years for
its light to reach us.
597
This redshift corresponds to a comoving or proper distance of 29 billion lightyears.
598

The size of the black hole is estimated at about 2x10
9
M. (An even further object is UDFj-39546284, a
compact galaxy of blue stars with unconfirmed z = ~10, corresponding to a comoving distance ~32 billion
lightyears.
599

The high luminosity of quasars we observe may be in part due to the directionality of the jets. The emission
may not be uniform for all angles, and the quasars that we observe may be those with jets pointing toward us.
Highly luminous quasars and AGNs overwhelm observation of the surrounding galaxy, but lower-luminosity
AGNs appear like galaxies (ZXI lecture 7).

Quasar Spectra
These exhibit a very broad energy distribution spanning 8 to 10 or 15 orders of magnitude, dominated at low
frequencies by the jet radiation. A turnover (flattening off of the rise) and infrared bump occurs in far
infrared. Peak power is radiated in the low-energy gamma rays. A prominent blue bump is often seen due to
excess of UV output (at 10
14
10
16
MHz), and quasars visually appear quite blue. There are often absorption
lines, often Doppler broadened and narrow lines... (see IMA2 p. 1098 for further details).
The following graph (in a 2001 paper for which ZXI is a co-author) is a composite of over 2200 SDSS spectra
of quasars with at least one broad emission line, adjusted to a rest frame. Broad absorption line quasars
(BAL quasars) are not included. The spectrum shows broad and narrow lines (the former are not seen in Type

595
http://ned.ipac.caltech.edu/level5/March02/Courvoisier/Cour3_5.html
596
UKIRT=United Kingdom Infrared Telescope (on Mauna Kea); UKIDSS = UKIRT Infrared Deep Sky Survey;
ULAS = UKIDSS Large Area Survey ULAS
597
http://www.eso.org/public/news/eso1122/
598
http://en.wikipedia.org/wiki/ULAS_J1120%2B0641
599
UDFj-39546284 distant compact galaxy:
http://www.astro.ucla.edu/~wright/CosmoCalc.html (Cosmology Calculator)
http://en.wikipedia.org/wiki/UDFj-39546284
2 AGNs, but there is some overlap). At wavelengths longer than the Ly emission, the continuum of the
geometric mean composite is well fitted by two power laws, with a break at 5000 . The frequency power-law
index, , is -0.44 from 1300 to 5000 and -2.45 redward of 5000 . The abrupt change in slope can be
accounted for partly by host-galaxy contamination at low redshift.
600

As with Seyferts and AGNs in general, it is believed that for Quasars of Type 1 that have broad emission lines,
we can see their high velocity gas (presumably in the accretion disk) whereas Type 2 quasars (which lack
broad emission lines) are tilted so that we do not see the high-velocity gas (or simply lack it (ZXI lecture 7).
Ultraluminous Infrared Galaxies
Many of these were found by the 1983 Infrared Astronomical Satellite (IRAS) and by later infrared-sensing
satellites. A Luminous Infrared Galaxy (LIRG), is a galactic body whose defining characteristic is that it
emits more than 10
11
L in the far-infrared. An even more luminous system, emitting more than 10
12
L in
the far-infrared, is called an Ultraluminous Infrared Galaxy (ULIRG). Most LIRGs and ULIRGs emit at least
90% of their light in the infrared. The infrared may arise from absorption and re-radiation by dust
enshrouding the radiation from an active galactic nucleus or quasar. These galaxies are cataloged in the IRAS
Revised Bright Galaxy Sample (RBGS).
601

High Cosmological Redshift of Quasars
As previously noted, high redshifts are seen with quasars, as high as 7.1 In the SDSS survey, currently
identifying more than 60,000 quasars, those in the top 1000 in redshift have z values ranging from 4.1 to
6.0.
602
There are more than 1000 having z > 4. A redshift of z=6.0 corresponds to an age at emission of 2.2

600
Vanden Berk DE, et al, Composite Quasar Spectra From The Sloan Digital Sky Survey, The Astronomical
Journal, 122:549-564, 2001 August.
601
602
http://cas.sdss.org/dr7/en/proj/advanced/quasars/query.asp
Note: for sorted output, add at the bottom of the suggested query order by z desc
Gigayear (i.e., the age relative to the Big Bang, which occurred 13.7 Gyr ago), a light travel time of 11.48 Gyr,
and a comoving radial distance of 21.07 Gly.
603

Redshifts can be caused by ordinary Doppler effects and by gravitational effects on light moving out of a
strong gravity field.
604
In addition, Cosmological Redshifts are caused by the expansion of space through
which the light travels. The total elongation of wavelength depends on how much expansion has occurred.
For such great distance, it is customary to quote z rather than an inferred distance, computed by wavelength
in the emitters rest frame compared to wavelength in the observers rest frame.

or
+
where R is the size of the universe (to be later defined). Thus a redshift of 7 means that the universe is 8
times larger than when the light was emitted (IMA2 p. 10991100).

Quasar Evolution
Quasars at large redshifts appear to us as they did many billions of years earlier. Equivalently, we see further
into the past when viewing quasars at increasingly large redshifts. Bright quasars were more common (had
greater space number density) in the distant past (large redshift). To remove the confounding effect of the
expansion of space in comparing density assessments, we define the comoving space density as:
coov sce es
e o oecs es e c

This is a way of saying the density the object will have at the current time, for which z = 0. As long as objects
are not being created or destroyed or evolving to other categories, etc., their comoving space density is
constant despite the expansion of space.
The comoving space density of quasars brighter than MB 25.9 is1000 times higher at z=2 than at z=0.
Quasar number density peaks at about z = 2.2, when the universe was about 1/4 of its present age, and that
95% of quasars are found between 5% and 50% of the current universes age (ZXI lecture 7).
605
However, this
appears to be due to a shift with time of the luminosity curves (Mg) rather than of the number of quasars
606

(see also IMA2 p. 1101 for diagram). Assessment of quasar populations also suggests that quasars on the
average tend to be dimmer for lower redshifts near 0 and thus greater proximity compared to quasars with
z=2.
But for 2 < z < 3, the comoving space densities for AGNs in X-ray wavelengths are seen (Silverman 2005) to
peak at about z = 2.5 and to falloff for higher z values:
607

603
http://www.astro.ucla.edu/~wright/CosmoCalc.html
604
http://en.wikipedia.org/wiki/Gravitational_redshift
605
http://www.astro.ucla.edu/~wright/CosmoCalc.html (to calculate age by given redshift)
606
Richards GT et al, The 2dF-SDSS LRG and QSO (2SLAQ) Survey: the z < 2.1 quasar luminosity function
from 5645 quasars to g= 21.85, Monthly Notices of the Royal Astronomical Society (MNRAS), Volume 360,
Issue 3, pp. 839-852 07/2005
607
Silverman JD, et al, Comoving Space Density of X-ray-selected Active Galactic Nuclei, ApJ 624 630, 2005

Perhaps the decreasing numbers of quasars at even higher z (z > 3) indicates that time for a growth phase of
SMBHs must elapse before quasars can develop.
The existence of a quasar may be limited to a galactic dynamical timescale. For all radio objects studied in a
HST study
608
(10 RLQs=radio-loud quasars, 13 RQQs=radio-quiet quasars, and 10 RGs=radio galaxies) for
which nuclear luminosities were brighter than absolute visual magnitude MV = 23.5, the host galaxies were
found to be massive ellipticals. (A few of the lower luminosity quasars were found to be in disk/bulge
systems.) Other than having an AGN, the ellipticals appeared indistinguishable from lower-z quiescent
galaxies found in cluster centers. All of the RLQs had SMBHs of at least 10
9
M, and the RQQs had M >
5x10
8
M. The greater abundance of RQQs compared to RLQs probably reflects the large size of the SMBH
required for a RLQ.
It is possible that the strongest quasar radio sources require not only large SMBHs but also rotation of the
SMBH (IMA2 p. 1104).
A 2004 SDSS-based study
609
found that for 12,698 quasars with 0.01 z 2.1, the majority of black holes
with mass > 10
8.5
M (and therefore capable of powering quasars) were in place by z 2. The quasar
bolometric luminosities steadily increase with z from z ~0.2 to z ~2.0. The ratios of Lbol / LEdd were no greater
than 1, suggesting that Eddington luminosity is still a relevant physical limit to the accretion rate of
luminous broad-line quasars at z 2. Regarding quasar lifetimes, McLure and Dunlop state, A fairly robust
limit on the lifetime of quasars with black hole masses 10
9.5
M is found be tQ > 2x10
8
years.

608
Dunlop JS, McLure RJ, et al, Quasars, their host galaxies and their central black holes, Monthly Notices
of the Royal Astronomical Society, Volume 340, Issue 4, 9 APR 2003
609
McLure RJ, Dunlop JS, Quasars, their host galaxies and their central black holes, Monthly Notices of the
Royal Astronomical Society, Volume 352 p. 1890, 2004
Timescale of AGN Variability
Many AGNs have luminosities of broad emission lines that can change by a factor of 2 in time intervals as
short as a few daysthis variability lags somewhat behind the variation in the continuum spectrum. (Narrow
lines do not exhibit nearly this degree of variability.) X-ray and visible output can fluctuate a few percent over
intervals as short as a few minutes. Some variations, such as the apparent magnitude of 3C 279, vary over
much longer intervals up to several years. (IMA2 p. 1104). See below regarding Blazars, BL Lacs and OVV
Quasars.
It is generally assumed that most, if not all, AGN are variable at some level. For example, virtually all of the
AGN observed in the Hubble Deep Field were determined to exhibit detectable variability over a period of two
years....
AGN Variability is an active area of research.

Polarization of Emissions
Quasar emission usually have low polarization of < 3% linear polarization. The core-dominant radio sources
(aka compact sources) are the exceptions, having up to 35% linear polarization. The lobe-dominant radio
sources have even higher polarization of up to 60% linear polarization (see for example 3C 47, shown on IMA2
p. 1106). AGN jets have high polarization of up to 40 to 50%.
Unlike thermal radiation, which is unpolarized, synchrotron radiation from relativistic electrons circling in a
magnetic field is polarized. This mechanism may account for some of the polarization seen with quasar
emissions.
In addition to linear polarization, circular polarization may be observed in jets. It can arise intrinsically in the
synchrotron process, or be produced secondarily by Faraday rotation:

Circular Polarization Distribution
610
(maximal in core) of Part of Jet of 3C 273
Circular polarization (color-scale given in mJy beam
-1
) superimposed on Stokes I contours.
The I contours for both sources start at 10 mJy beam
-1
. Image date August 28, 2003.
The additional phase calibration [for this image] has corrected the phase error and reveals
that the real circular polarization in the source is on the jet core and not on the jet
component...

Fanaroff-Riley Luminosity Classes
It was first noticed by B.L. Fanaroff and J.M. Riley (1974) that the relative positions of regions of high and
low surface brightness in the lobes of extragalactic radio sources are correlated with their radio luminosity.
This conclusion was based on a set of 57 radio galaxies and quasars... Fanaroff and Riley divided this sample
into two classes using the ratio RFR of the distance between the regions of highest surface brightness on
opposite sides of the central galaxy or quasar, to the total extent of the source up to the lowest brightness
contour in the map. Sources with RFR < 0.5 were placed in Class I [like NGC 1265 and 3C 449], and sources
with RFR > 0.5 in Class II [like Cyg A, 3C 47, and most quasars]. It was found that nearly all sources with
luminosity L(178 MHz) < ~2 x 10
25
h100
-2
W Hz
-1
str
-1
were of Class I, while the brighter sources were nearly all of
Class II... [IMA2 specifies this distinction for 1.4 GHz]
Fanaroff-Riley Class I (FR-I): Sources in this class have their low brightness regions further from the central
galaxy or quasar than their high brightness regions. The sources become fainter as one approaches the outer
extremities of the lobes and the spectra here are the steepest, indicating that the radiating particles have aged
the most. Jets are detected in 80 per cent of FR-I galaxies.... FR-I sources are associated with bright, large
galaxies (D or cD) that have a flatter light distribution than an average elliptical galaxy and are often located
in rich clusters with extreme X-ray emitting gas...
Fanaroff-Riley Class II (FR-II): This class comprises luminous radio sources with hotspots in their lobes at
distances from the center which are such that RFR > 0.5. These sources are called edge-darkened, which was
particularly apt terminology when the angular resolution and dynamic range used in observing the classical
sources was not always good enough to reveal the hotspots as distinct structures. In keeping with the overall
high luminosity of this type of source, the cores and jets in them are also brighter than those in FR-I galaxies
in absolute terms; but relative to the lobes these features are much fainter in FR-II galaxies. Jets are detected
in < 10 per cent of luminous radio galaxies, but in nearly all quasars. The jets have small opening angles (< 4
deg) and are knotty; the jet magnetic field is predominantly parallel to the jet axis except in the knots, where
the perpendicular component is dominant... The most striking feature of the jets in the FR-II class is that
they are often one-sided... FR-II sources are generally associated with galaxies that appear normal, except
that they have nuclear and extended emission line regions. The galaxies are giant ellipticals, but not first-
ranked cluster galaxies...
611

Blazars: BL Lacs and OVV Quasars
Some AGNs known as BL Lacs (also referred to as blazars) are among the most active variable sources
known in the universe.
612

BL Lacs are named after the most well-known object, BL Lacertae or BL Lac.
Blazars are thought to be active galactic nuclei, with relativistic jets oriented close to the line of sight with the
observer.
613
Collimated beams exhibiting slight fluctuation in the direction of the beam could cause marked
variability in the radiation reaching the observer. BL Lacs have a featureless continuum, and rapidly vary
over hours to days or months in brightness. They are strongly polarized and have few emission lines.

610
Homan DC et al, MOJAVE: Monitoring of jets in active galactic nuclei with VLBA experiments. II. First-
epoch 15 GHz circular polarization results, The Astronomical Journal, 2006, available at
http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=1681&context=physics_articles

611
http://ned.ipac.caltech.edu/level5/Glossary/Essay_fanaroff.html
612
http://gtn.sonoma.edu/resources/agn_variability/agn_variability.php
613
http://en.wikipedia.org/wiki/Blazar
Redshifts where observed are high, indicating that they are at cosmological distances. Most are in elliptical
galaxies.
Another subdivision of blazars are the Optically Violent Variable quasars (OVV). These are like BL Lacs
except that they are much more luminous and their spectra may display broad emission lines.

LINERS
These are Low Ionization Nuclear Emission-line Region galaxies: They have very weak or no continuum
(low luminosities); but fairly strong emission lines of low-ionization species. They are similar to the low limit
of the range of luminosity for Seyfert 2 class galaxies, and are detectable in starburst galaxies. It is unclear if
they represent the low-luminosity limit of AGNs. (IMA2 p. 1107)

A Unified Model of Active Galactic Nuclei (including Quasars)

It is reasonable to pursue an attempt to simplify and unify the seemingly disparate types of entities with
AGNs. For one, the Luminosity LH of AGNs in the H line has been shown to be proportional to LFC, the
luminosity of the featureless continuum at 480 nm, for quasars, Seyferts, and NLRGs (IMA2 p. 1109). This
suggests that the ionized H atoms are generated by photoionization by the continuum radiation. There is also
a line of evidence that suggests that within Seyfert 2s are Seyfert 1 nuclei that are hidden by optically thick
material, and that the light we see arrives indirectly by reflection.

Size of the AGN
The rapid time variability is the strongest clue as to the scale of the AGN. The maximum radius R of an object
that can brighten throughout over a period of t is given by R = ct/, where is the Lorentz factor

v

This is of course based on speed of light and special relativity considerations. For t=1 hour, R = 7.2 AU, a
comparatively very small size for an object of such intense luminosity as an AGN. The Eddington limit sets an
upper limit on luminosity, LEdd, such that LAGN < LEdd. The lower limit for the AGN mass becomes MAGN > 3.3
x 10
8
M. The upper limit set by the Schwarzschild radius is M = 3.7 x 10
8
M. The close convergence of
these values supports that idea that a SMBH is present in the center.

Luminosity Through Accretion
The most efficient way to convert gravitational potential energy to other forms (heat, light) is by mass
accretion. Accretion of matter onto a neutron star of 1.4 M releases 21% of the rest energy of the matter.
With a BH, there is no surface to strike, the matter will just disappear through the event horizon. However,
matter spiraling in decreasing orbit toward a black hole will pile up on an accretion disk, allowing viscous
dissipation of PE and KE as heat and radiation. The smallest stable orbit is at 3RS, where RS is the
Schwarzschild radius. To reach this point, the matter must give up 5.2% of its rest mass in binding energy
(the energy needed to remove it to infinity). For rapidly rotating black holes, the smallest stable orbit is at
prograde 0.5 RS (or retrograde 4.5 RS), yielding up 42.3% of rest mass energy! The accretion of matter
through a disk around a rapidly rotating black hole is an extremely efficient way of producing large amounts
of energy (IMA2 p. 1111). The matter at the inner edge of the disk rotates at the same speed as the BH. The
RS increases for rising MBH, so that characteristic disk temperature decreases for increasing MBH, given by
Tdisk M
-1/4
. According to an example shown, a typical disk temperature could be 73000 K, which by Wien
Displacement places the peak emission in the extreme UV, 39.7 nm (IMA2 p. 1112). More sophisticated
calculations yield a disk temp of several hundred thousand K. The big blue bump would be a thermal
signature from the UV excess arising at the disk.

The Accretion Disk, Torus, and Size of the SMBH
These have been extensively modeled, and this can only be a limited summary. According to the unified
model, the different appearances are due to differing viewing angles (in addition to differences in size and age,
etc.) Here is one sample representation, though simplified compared to the textbook description and drawing:

Unified Model of AGN with Black Hole and Accretion Disk.
614

614
modified from www.ifa.hawaii.edu/~treister/talks/ifa08.pptx

Unified Model of AGNs (Annotated, from NASA)
615

If the disk overheats, it puffs up into an ion torus, supported by pressure of the hot ions. If the luminosity
becomes super-Eddington, the radiation pressure lifts the matter into a more inflated radiation torus.
The disk may have several sections. The BH is in the center, unseen and tiny compared to the size of the
disk. Innermost in the disk (starting at < 1000 RS) is a thick hot disk where radiation pressure exceeds the
inward forces and the surface expands locally. This gives a concave surface to the mid disk that can be
radiated by the thick hot disk. Beyond 100,000 RS, the disk breaks into clouds.
The black holes Schwarzschild or Kerr
616
radius is tiny compared to that of the torus. For a SMBH mass of
10
9
M, the Schwartzschild radius is 3x10
12
m or about 0.1 light day or (1/2 this for Kerr equatorial radius
for maximal rotationIMA2 p. 641).
The radius of the torus is hard to pin down. There are few images which resolve the core of an AGN. The
putative torus may be perhaps 200 to 400 light years in radius (400 from 1/2 of diameter in HST image
617
of
NGC4261, 228 from IMA2 p. 1121, the latter giving a ratio of torus radius to Schwarzschild radius of nearly
8 orders of magnitude).
The HST image of nearby galaxy NGC 4261 shows a 800 light-year-wide [400 ly radius] spiral-shaped disk of
dust fueling [the SMBH],
618
but the text does not use the word torus. IMA2 p. 1121 however describes this
same image as the core of an elliptical radio galaxy and optical LINER, showing a bright nucleus surrounded
by a large, obscuring torus whose radius measures about 70 pc (228 ly). The SMBH is estimated to have a
mass of 5 x 10
8
M.
619

615
http://www.auger.org/news/PRagn/about_AGN.html
616
http://en.wikipedia.org/wiki/Kerr_metric
617
http://www.mssl.ucl.ac.uk/www_astro/agn/agn_quasartour.html
618
http://hubblesite.org/newscenter/archive/releases/1995/47/text/
619
NGC 4261:
http://blackholes.stardate.org/directory/factsheet.php?p=NGC-4261
http://www.stsci.edu/stsci/meetings/shst2/ferraresel.html
For a maximally rotating 10
7
M SMBH, the estimated RKerr = RS/2 = 1.61x 10
-6
lightyears. The estimated
ratio of Torus radius (228 lightyears) to RKerr of the SMBH is roughly 1.5 x 10
8
, similar to the estimate in
IMA2.

HST image of core at center of NGC4261, an elliptical radio galaxy and LINER,1995
including an 800 light-year-wide [400 ly radius] spiral-shaped disk of dust.

Mechanism for Producing a Relativistic Outflow of Charged Particles
The rotating conducting accretion disk can generate a magnetic field. The varying magnetic field may
generate a large electric field near the surface of the disk, which causes acceleration of charged particles to
relativistic speeds.
Relativistic jets may provide evidence for the reality of frame-dragging. Gravitomagnetic forces produced by
the LenseThirring effect (frame dragging) within the ergosphere of rotating black holes combined with the
energy extraction mechanism by Penrose have been used to explain the observed properties of relativistic
jets.
620

The relativistic outflow may also be produced by the Blandford-Znajek effect. With it, the accretion disk can
tap the rotational energy of the BH itself. This is the most popular theory for the extraction of energy from
the central black hole. The magnetic fields around the accretion disk are dragged by the spin of the black
hole. The relativistic material in the jets is possibly launched by the tightening of the field lines.
621
The BH
rotating in a magnetic field (of perhaps 1 T) can be thought of as an electrical conductor and generates an
electrical potential between its poles and equator across a resistance of about 30 . A current flows, the BH
spin is reduced, and power is generated that is imparted to the outflowing particles (paired electrons and
positrons) and to EM radiation. Up to 9.2% of the rest energy of a maximally rotating BH can be extracted in
this manner. The electrons spiraling about the magnetic field lines emit synchrotron radiation that

620
http://en.wikipedia.org/wiki/Frame-dragging
621
http://en.wikipedia.org/wiki/Relativistic_jet
contributes to the continuous spectrum of AGNs. This synchrotron radiation spectrum follows a power law
(IMA2 p. 1115).

X-ray Generation
These are formed in several ways in addition to synchrotron radiation. The soft lower-energy X-rays may
arise from the high-frequency end of the accretion disk spectrum. It is also possible that higher energy X-rays
are produced by inverse Compton scattering of photons off the relativistic electrons
622
this mechanism can
produce very high energy photons, having final energy (from headon collisions) of up to
2
0. Thermal
bremsstrahlung scattering can also produce X-ray energy emissions.

Broad Line and Narrow Line Emission
These are the result of photoionization by the continuum radiation. The broad lines arise from permitted
atomic transitions. The narrow lines may also involve forbidden transitions. Their variabilities differ, and
these probably arise from different regions under different conditions.
The broad line region is close to the center of the disk. The broad-line emissions can respond rapidly to
changes in the continuum radiation in as short as 1 week... They are clumpy, and contain partially ionized
gas of about 10
15
m
-3
to 10
16
m
-3
electron number density. The temperature is about 10
4
K.
The narrow-line regions have much lower electron density, about 10
10
m
-3
, but have large overall mass and a
temperature of about 10
4
K [same as broad-line region], and for these reasons can exhibit forbidden as well as
allowed transitions. (IMA2 p. 1116-1121)

Supermassive Black Hole Masses in AGNs versus Luminosities
These can be estimated using emissions from broad-line regions. Assuming a width and a corresponding
rotational velocity of the lines (e.g., 5000 km s
-1
)
623
and assuming a radius r = 10
15
m, the black hole mass
may be estimated as
MBH = rv
2
/G = 1.9 x 10
8
R..
Another technique uses reverberation mapping, in which the lag time (time delay ) between changes in
brightness of the continuum lines vs. the emission lines is evaluated. This technique may be most useful at
high z and has the advantage of not requiring spatially resolving the central region.
The masses of SMBHs by reverberation mapping have been estimated at 0.2x10
7
M to about 10
9
M.
The bolometric luminosities of SMBHs in 234 AGNs have been shown in the following graph
624
to only roughly
correlate with their masses (see also somewhat different graph at IMA2 p. 1118, which used by reverberation
mapping).

622
http://www.cv.nrao.edu/course/astr534/InverseCompton.html
623
Doppler widths in the range 1000 to 25,000kms
1
per http://www.astronomy.ohio-
state.edu/~peterson/Docs/chapter_3.pdf
624
Woo Jong-Hak, C. Megan Urry, Active Galactic Nucleus Black Hole Masses and Bolometric Luminosities,
ApJ, 579:530-544, 2002 November 10,
http://iopscience.iop.org/0004-637X/579/2/530/pdf/0004-637X_579_2_530.pdf

Bolometric luminosity vs. black hole mass for 234 AGNs [using masses that were derived from optical
luminosity and broad-line velocity]. There is little if any correlation... For a given black hole mass,
there is a large range of bolometric luminosities, spanning 3 or more orders of magnitude. The Eddington
limit [upper dashed line L=LEdd] defines an approximate upper limit to the luminosity,
but the absence of objects from the lower right of the diagram (low-luminosity, high-mass AGNs)
is a selection effect. For example, this part of the diagram would be occupied by BL Lac objects
and low-luminosity radio galaxies... The rectangle delineates where BL Lacs are expected,
but the authors note that for BL Lacs, we do not have good estimates of bolometric luminosity.
(Woo and Urry, 2002)
625

Summary of the Unified Model, With Emphasis on the Observers Angle of View
According to the unified model, the central engine of an AGN is a supermassive black hole with a surrounding
accretion disk. It is powered by conversion of gravitational potential energy to rotational kinetic energy,
thermal energy, and synchrotron radiation. The structure of the disk (e.g., how puffed up it is) depends on the
ratio of accretion disk luminosity and the Eddington luminosity limit (which is mass-dependent). The
accretion rates typically are 1 to 10 M yr
-1
. The black holes Schwarzschild radius or Kerr
626
radius is tiny
compared to that of the torus (see details above). For a SMBH mass of 10
9
M, the Schwartzschild radius is

625
ibid.
626
http://en.wikipedia.org/wiki/Kerr_metric
about 0.1 light day (1/2 this as Kerr equatorial radius for maximal rotation). The torus radius may be
perhaps 200 to 400 light years. The jets reach out some 15 kpc from the nucleus.
The following appearances based on view perspective are postulated (see diagrams):
Seyfert 2, NLRG: seeing the disk from the side is blocked by the obscuring torus so cant see the AGN
or its radiation directly, and cant see Broad-line Region contained inside the torus. The Narrow Line
Region is positioned further out, so its emissions are not obscured by the torus.
Seyfert 1, BLRG, QSO: viewing from about halfway between torus and central axis, sometimes on the
no/weak jet side, can see the AGN partially, greater luminosity. Can see Broad-line region + Narrow
Line Region. Radio is weak w Seyfert 1 & QSO, strong with BLRG.
Blazars (BL Lac, OVV), Radio Loud Quasars QSR: viewing close to (for QSRs) or directly on (for
Blazars) the strongest jet axis. Strong radio (esp. for Blazars), strong continuum, broad + narrow
emission lines.
Radio Quiet Quasars QSO: Can also arise from viewing from the no/weak jet side.

AGN Evolution
Merger with other galaxies and companions help to provide fuel to the accretion disk of an Active Galactic
Nucleus (IMA2 p. 1129). As an AGN runs out of fuel, the central luminosity may decline before the radio lobes
exhibit decreasing emission. AGNs may undergo periods of quiescence punctuated by reactivation from
mergers, etc. The lifetimes of AGNs are not known.

Radio Lobes and J ets

Radio-loud galaxies also tend to be strong emitters in X-ray.

Generation of Jets and Radio Lobes
The jets consist of charged particles that have been accelerated away from the nucleus to relativistic speeds in
opposite directions. IMA2 states that the jets may be electrons and positrons or electrons and ions. The
magnetic field of the disk becomes frozen into the plasma of the jets of outgoing charged particles. The
remarkable collimation of some jets suggests that the process takes place close to the central engine powering
the jet. The collimation may arise from the hot accretion disk around the SMBH. Material will pile up on the
inner centrifugal barrier, leaving an empty nozzle more centrally through which the jet may flow. Poorly
understood magnetohydrodynamic MHD effects may also play a role.
A jet encounters resistance in the interstellar medium within the galaxy and in the intergalactic medium
beyond. A shock front develops and the jets slowssome of the energy splashes back and also becomes
disordered. The highly turbulent nature of this disorder is quite difficult to model.
A single lobe (such as Cygnus A) may hold 10
53
to 10
54
joules as kinetic and magnetic energy (the latter given
by M=B
2
/20 in units of J m
-3
). Given this store of energy and the observed luminosity rates, it is possible
that a radio lobe might have a lifetime of more than 10
8
years. IMA2 p. 1125 estimates the magnetic field in a
radio lobe at 41 nT (range about 1 nT to 100 nT).
The jet radiation emissions exhibit a power law spectrum and strong linear polarization, consistent with origin
from synchrotron radiation. Particles may be accelerated by magnetic squeezing by shock waves, and also
radiation pressure.

Relativistic Beaming
Jets often exhibit relativistic beaming or headlight effect, another consequence of special relativity,
specifically the Lorentz transformation of velocities. Rays of light emitted by a moving object appear to a
relatively nonmoving observer as if they have arisen concentrated into a forward angle more in the direction of
the object's motion. Thus light emitted by a moving object appears to be compressed or concentrated into a
cone about its direction of motion, an effect called relativistic beaming, and this can intensify the brightness
of the beam. This concentration of light in the object's forward direction is also referred to as the headlight
effect. (IMA2 p. 101-2) This contributes to the appearance of one-sided beams when opposing beams are
more likely.

Apparently Superluminal Jets
Some jets appear to have superluminal velocity exceeding c. For example, 3C 273 exhibits knots (blobs) with
an angular velocity of = 0.0008 yr
-1
. The distance to 3C 273 is 440h
-1
Mpc. Then assuming h = hWMAP, the
apparent transverse velocity = vapp = d = 7.85c. The explanation involves special circumstances in which the
blob has been ejected at relativistic speed so that a narrow angle is formed between the jet direction and the
observers line of sight. This phenomenon is caused because the jets are travelling very near the speed of
light and at a very small angle towards the observer. Because at every point of their path the high-velocity
jets are emitting light, the light they emit does not approach the observer much more quickly than the jet
itself. To be more clear, the jet is essentially chasing the light it emits. This causes the light emitted over
hundreds of years of travel to not have hundreds of lightyears of distance between it, the light thus arrives at
the observer over a much smaller time period (ten or twenty years) giving the illusion of faster than light
travel.
627
The maximal apparent transverse velocity depends on the angle and the actual velocity of the jet.
The formula is

where T = vT/c (the apparent transverse velocity divided by c), and = Lorentz factor. If >> 1 and the jet
velocity v is close to c, then
.

Using Quasars to Probe the Universe

Gravitational Lensing
QSRs are among the most distant objects known.
Fritz Zwicky (1898-1974) suggested that galaxies
could cause gravitational lensing. Many examples
have been found. The image of a QSR lensed by a
galaxy that lies exactly on the line of sight will be an
Einstein Ring.
628
If the lensing object is off axis, the
ring will appear as a partial ring or arc.
If the lens is extended and elliptical in shape, 3 or 5
images can be formed, for instance, the Einstein
cross (G2237+0305 = QSO 2237+0305).
629

Apparently there are at least two other Einstein
crosses, but I have not yet located information on
them.
630

Image to right is of the Einstein Cross Gravitational
Lens. This is a gravitationally lensed quasar that
sits directly behind galaxy ZW 2237+030. Four
images of the same distant quasar appear around
the foreground galaxy (the latter probably obscures a
5th quasar image) due to strong gravitational
lensing. The angular separation between the upper

627
http://en.wikipedia.org/wiki/Superluminal_jet
628
http://en.wikipedia.org/wiki/Einstein_ring
629
http://en.wikipedia.org/wiki/Einstein_Cross
630
http://scienceworld.wolfram.com/physics/EinsteinCross.html
and lower images is 1.6 arcseconds. (HST using ESA Faint Object Camera)
631

Light from distant galaxies or quasars passing through galaxy clusters (such as Abell 1689 or Abell 2218) may
be distorted into a series of arcs and other deformed shapes.
632
The Einstein Circle or Einstein Ring is a
special case of gravitational lensing, caused by the exact alignment of the source, lens and observer (which
almost never occurs). This results in a symmetric ring around the lens. The angle of the ring in radians is
given by the Einstein Radius E:
633

) (
)
where DLS = Angular Diameter Distance from Lens to Source, DL = Angular Diameter Distance from lens to
observer, and DS = Angular Diameter Distance from source to observer. Angular Diameter Distance is a
relativityrelated measure. Note that for nearer objects, DLS = DS DL, but at cosmological distances,
DLS <> DS DL due to relativity effects. If the source is off-axis, multiple arcs or images may be seen rather
than a ring.
The multiple lensed images of a quasar may vary in relative intensity with time. In addition, the images are
taking different paths in space so that their times of arrival for the same source time may differ by more than
a year, further complicating things when the source is variable.
The differing delay among multiple images has been used as an independent estimate of the Hubble constant
H0 (69 vs. 71 km s
-1
Mpc
-1
by WMAP), using for instance the Twin Quasar,
634
for which the distant quasar is at
z = 1.4. The two quasar images are separated by 6 arcsec, and there is a 417 day time lag between the two
images.
635

Gravitational Microlensing is proving useful, for example in the Microlensing Planet Search Project (MPS):
Planetary systems in the foreground of the Galactic Bulge form a class of (planetary) binary lens systems that
can be detected through photometric microlensing experiments. An Earth mass planet orbiting a lensing star
can perturb the would-be single lens microlensing light curve due to the lensing star, briefly but
spectacularly. Gravitational microlensing is the only known ground-based method to probe earth mass
planets orbiting around main sequence stars. (Earth-mass objects orbiting a neutron star were discovered in
1992 by Wolszczan and Frail.)
636

David Darling states: When aligned with a more distant star, a microlens deflects, distorts, and brightens
the image of the background source in a way that is sensitive to the strength and geometry of the lensing
star's gravitational field. If the lensing star is orbited by a planet then the result can be a microlensing
anomaly known as a binary lens, one of the characteristics of which is the appearance of so-called caustic
curves in the plane of the source. As the source sweeps across the caustic curves, due to the relative motion
of source and lens, sharp peaks appear in the light curve. Planets located at a distance between about 0.5
and 1.5 times the Einstein radius of their host star are said to be in the "lensing zone," since they are most
likely to produce, with the help of the host, a sizable caustic structure. This distance is typically a few
astronomical units, or approximately the distance of Jupiter from the Sun. Calculations (for example, by
Bohdan Paczynski) suggest that microlensing anomalies due to Jupiter-mass planets in Jupiter-like orbits
will last about 1 to 3 days and be detectable in 10 to 30 percent of cases where such planets are in orbit
around a lensing star that is located roughly halfway between us and the center of the Galaxy. Smaller mass
planets will create smaller caustic structures which (1) reduces the chance of the background source passing
directly behind the anomaly and (2), in cases where the source does pass behind the anomaly, results in a
shorter and less sharp increase in brightness. The probability of detection and the duration of the lensing
peak decrease as the square root of the planet's mass. So, an Earth-mass planet (320 times less massive than
Jupiter), orbiting at about the radius of its host's Einstein ring, will give rise to a caustic structure 18 times

631
632
http://en.wikipedia.org/wiki/Gravitational_lens
633
http://en.wikipedia.org/wiki/Einstein_radius
634
Twin Quasar: (Q0957+561) or (QSO 0957+561 A/B) or (SBS 0957+561)
635
http://en.wikipedia.org/wiki/Twin_Quasar
636
http://bustard.phys.nd.edu/MPS/
smaller than that of a Jupiter-mass planet in a similar orbit, and any light curve variations caused by it will
last for only a few hours rather than a few days.
637

To emphasize, microlensing is manifested not by image splitting, distortion, or angular magnification, but by
a smooth brightening (loosely spoken of as magnification) and dimming of the lensed source relative to its
unlensed state. (Image that follows taken from here
638
):

The search for exoplanets using this state-of-the art technique and other research techniques warrants much
more attention when I find the time.

The Lyman-Alpha Forest
The resting Lyman-alpha line
639
is at 1215.668 Angstroms or 121.5668 nm, thus in the vacuum UV. High
redshift quasars usually emit a prominent Lyman- emission line, which is strongly redshifted by the time it
reaches us (IMA2 shows a Lyman- emission line for QSO 1215+333 at about 440 nm). When broad-
spectrum light from the distant quasar passes through multiple H I clouds of neutral hydrogen, absorption
transitions occur at the Lyman- wavelength at each location. These create absorption lines that are then
redshifted as the light propagates toward the observer, with greater redshift in absorption lines occurring for
clouds closer to the source. Each individual cloud leaves its own spike-like Lyman- absorption line at a
different position in the observed spectrum, and the numerous and variably redshifted Lyman- absorption
lines taken as a group are termed the Lyman-Alpha Forest.

637
http://www.daviddarling.info/encyclopedia/M/microlensing.html
638
http://en.wikipedia.org/wiki/Gravitational_microlensing
639
http://en.wikipedia.org/wiki/Lyman-alpha_line

Cartoon showing single maximally redshifted Lyman- emission peak (red on the right),
and a few of the often many Lyman-Alpha Forest absorption lines (various colors) arising
from clouds positioned at varying redshift distances between source quasar and observer.
Absorption lines closer to the observer are less redshifted and depicted as blue or green here.
640

Spectra of nearby quasar 3C 273 (z = 0.158) and of a distant quasar Q1422+2309 (z = 3.62).
The spectra are scaled to a common scale in emitted wavelength so that
the Lyman-alpha emission peak is shown at the same wavelength, ~1216 , in both cases.
The Lyman-alpha forest of hundreds of absorption lines for the more distant quasar
is apparent at the lower left. It has been a long journey!
641

640
http://www.astro.ucla.edu/~wright/Lyman-alpha-forest.html , Edward Wright 2004.
641
modified slightly by MCM from http://www.astr.ua.edu/keel/agn/forest.gif, by Bill Keel

Ionized Metal Absorption Lines in Quasars
The narrow absorption lines of ionized metals seen in ground-based quasar spectra vary by z. For z < 1.5, the
Mg II lines dominate in ground based spectra (and are accompanied by Si II, C II, Fe II, and ASl II lines. For
1.2 < z < 3.5, however, C IV, Si IV, N V, and O IV lines are commonly seen. The latter may be produced by
strong ionization of clouds by OB stars in young galaxies. These clouds also have lower abundances of heavy
elements, consistent with their arising in young galaxies.

Intergalactic Clouds
The comoving space density of intergalactic clouds appears to have been substantially greater in the past.
They are not grouped in clusters but appear to be distributed randomly.
Interesting simulations by Renyue Cen of spatial density of clouds of H I may be seen here.
642
(The
simulations also include many other parameters, apparently in a comoving cubical volume 25-Mpc on a side,
although the details are beyond my skill level) Look for instance at H I density at z = 6 down to z = 0, a
sequence clearly showing the decline of density with lower z (younger age).
Incidentally, Cen also presents the following elegant simulation
643
of galaxy formation, set to a lovely O Mio
Babbino Caro soundtrack.

642
http://www.astro.princeton.edu/~cen/PROJECTS/p2/p2.html
643
http://www.astro.princeton.edu/~cen/AMR_GalaxyFormation-2.mp4
Cosmology (Chapter 29, omitted)
The Early Universe (Chapter 30, omitted)

(omitted, insufficient time, maybe next year)

Astrophysics 2012

Uploaded by

Copyright:

Available Formats

Astrophysics 2012

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Astrophysics 2012

Uploaded by

Copyright:

Available Formats

What are some basic astronomical terms and concepts discussed?

What are some basic astronomical terms and concepts discussed?

What are the different distance measures used in astronomy?

What are the different distance measures used in astronomy?

Page 1 of 249 Astrophysics_ASTR322_323_MCM_2012.

docx 29 Jun 2012

(in the Blue-Band B)

= 5172 kg/W (calculated with above values).

), is not the same as astronomical He I or

apply to main sequence [luminosity class V] stars only

is the amount of energy liberated per kg of material, and

are the mass fractions of the two

e) and emissions of positrons, neutrinos, and gamma rays (IMA2 p. 309).

, with emission of electrons, neutrinos, and gamma

ions (a hydrogen atom with

, the authors derive A = 1.086, so that the

, Surface Acceleration , and Ratios of P/P

, so that gas moves

). For higher overtones, the disproportion between surface versus deep

(sin and/or cos terms multiplied by complex exponentials in

), both of which are highly luminous (see

, namely the emissions at 1238

(t1/2 = 6.1 days), the rate of decline slows

. The curve now follows the decay curve of o

(t1/2 = 77.2 days).

(t1/2 = 272 days). See the article for

/c)are entirely inadequate to support a WD against complete collapse. The

were to remain constant, the characteristic lifetime would be the time

diagram (see graph on left).

. However, when perturbed, instead of

for a semi-detached binary with mass equaling its companion is roughly

is the rate of matter infalling, = the Stefan

negative, thus cause the RH

is the frequency (yr

/ (the ratio of post-SN M1mass to total initial binary mass 1/2),

, where the bar

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.