Detecting High-Z Galaxies in The Near Infrared Background: Bin Yue, Andrea Ferrara, K Ari Helgason
Detecting High-Z Galaxies in The Near Infrared Background: Bin Yue, Andrea Ferrara, K Ari Helgason
Detecting High-Z Galaxies in The Near Infrared Background: Bin Yue, Andrea Ferrara, K Ari Helgason
2 October 2015
ABSTRACT
INTRODUCTION
The near-infrared background (NIRB) contains a considerable fraction of the collective radiation emitted by stars
in galaxies through cosmic times. As such, it offers a
unique opportunity to study faint high-z galaxies that
remain largely undetected in deep galaxy surveys (see
e.g. Salvaterra & Ferrara 2006; Fernandez & Komatsu 2006;
Fernandez et al. 2010, 2012, 2013; Fernandez & Zaroubi
2013). This is particularly important, as these objects are
commonly believed to provide most of the ionizing power
to drive cosmic reionization (Choudhury & Ferrara 2007;
Raicevic et al. 2011; Salvaterra et al. 2011). NIRB might
also help characterizing the stellar populations of the first
cosmic systems (Salvaterra & Ferrara 2003; Salvaterra et al.
2006; Santos et al. 2002; Kashlinsky et al. 2002, 2004,
2005; Magliocchetti et al. 2003; Cooray & Yoshida 2004;
Cooray et al. 2004). The most recent studies have converged
on the prediction that on scales of 1000 the fluctuation
>
level from galaxies at z
5 is 103 nWm2 sr1 at 3.6 m
(Cooray et al. 2012a; Yue et al. 2013a; Helgason et al.
2015).
However, extracting such signal from available data
has been so far very challenging. Even when the deepest
galaxy subtraction from NIRB maps is applied, the domi-
Yue et al.
2.1
High-z galaxies
(MUV
, z)dMUV
=
dMh ,
(1)
dM
MUV
Mh
h
where is the UV LF at 1600
A. For this, we use
the Schechter parameterization with the redshift-dependent
fitting parameters given in Bouwens et al. (2015). As a
reference, our minimum mass 5 108 M , corresponds
at z = 5, 8, 10 to an absolute magnitude MUV =
10.5, 12.1, 13.0, respectively. Luminosity at other UV
wavelengths is obtained through the luminosity-dependent
Spectral Energy Distribution (SED) slope (i.e., f )
in Bouwens et al. (2014)3 . However, generally speaking this
<
power-law only holds at
20003000
A, while we need luminosities at least until 4.5/(1+z) m, say 7500
A when z =
5. Therefore at >2000
A we use the SED template from
Starburst994 (Leitherer et al. 1999; V
azquez & Leitherer
2005; Leitherer et al. 2010), adopting a continuous star formation mode, with metallicity 0.1 Z and 200 Myr stellar
age. The SB99 SEDs are normalized to match the power-law
form at 2000
A.
The flux received in each pixel in the map is the sum of
radiation from all galaxies seen by the pixel,
F (0 ) = 0
X Lj ()(1 + zj )
,
(pix )2 j 4rj2 (1 + zj )2
1
(2)
http://homepage.sns.it/mesinger/DexM___21cmFAST.html
d
= 19.5 + dM
(MUV + 19.5), the values of 19.5 and
UV
d
dMUV
(3)
2.2
In addition to the flux from high-z (z > 5) galaxies, the observed NIRB also contains radiation from unresolved, lowz galaxies, and an excess radiation from unknown sources
(Yue et al. 2013b, 2014; Cooray et al. 2012b; Zemcov et al.
2014). We collectively refer to these two components as contamination, since in this work the targeted signal is the flux
from high-z galaxies.
We generate random maps to model this contamination. The contamination maps have mean flux 1.0 (0.7)
nWm2 sr1 at 3.6 (4.5) m. The flux fluctuations reproduce the sum of (i) the angular power spectrum of
the power excess (see Yue et al. 2013b) matching available
observations (Cooray et al. 2012b; Kashlinsky et al. 2012),
and (ii) the angular power spectrum of low-z galaxies
(Helgason et al. 2012) producing shot noise level matching
Kashlinsky et al. 2012 (4.8 1011 nW2 m4 sr1 at 3.6 m
and 2.2 1011 nW2 m4 sr1 at 4.5 m, the corresponding
subtraction magnitude is 25). Contamination maps are
constructed as follows:
A white noise map, i.e. a Gaussian random field, is generated.
This map is then transformed into frequency space by
FFT.
For each complex p
number in frequency space, its modulus is rescaled to be P (q), where P is the given power
spectrum and q is the spatial frequency. The zero-frequency
(q = 0) element is set to be the mean flux.
The above map is then transformed back into real space
by inverse FFT, resulting in a synthetic image with the same
2-point clustering properties as the measured P (q).
In Fig. 1 we plot a single realization of the contamination
map at 3.6 m as an example (bottom right). The contamination is not correlated with the high-z galaxy component;
however, it adds noise to the cross-correlation signal. To account for the statistical variance of the contamination, we
make 30 independent realizations of the maps. In Fig. 2 we
6
3
3.1
h(F1.6 F0 ) i
.
h(F1.6 )2 i h(F0 )2 i
(4)
Yue et al.
Figure 1. Upper: The 1.6 m flux map constructed from resolved LBGs with Hlim = 25 (left) and Hlim = 27 (right) respectively. Lower:
Map of the 3.6 m flux from galaxies with 5 < z < 10 (left) and contamination map at 3.6 m (right). The mean flux is 1 nWm2 sr1 .
(0.3)2 and (1.2)2 deg2 , representing a survey region similar to HUDF/XDF, UDS, and an hypothetical larger field,
respectively.
Before calculating the correlation coefficient, in both
maps we mask the pixels containing galaxies brighter than
25 at either 3.6 m or 4.5 m. From this procedure we obtain
the source-subtracted NIRB map. The correlation coefficient
vs. limiting LBG magnitude is shown in Fig. 3. The filled
regions are the 1 variance of all sub-maps with the same
area however cut out from different parts of the full map.
Note that for each signal map we have 30 contamination
realizations, so even for the (1.2)2 deg2 case we have 120
samples.
Fig. 3 shows that, it is indeed feasible to detect the correlation from the mock maps, even from a relatively shallow
survey with Hlim 25, which is 0.04. By pushing the
limiting magnitude fainter the correlation coefficient rapidly
increases and becomes a factor 2 higher, and then approaches 0.09 more slowly towards Hlim 29. It is worth
noting that in small area fields the measured correlation coefficient has 30 80% relative field-by-field scatters when
>
Hlim
26, and even larger scatters when Hlim < 26. In some
cases there would be no cross-correlation detected, due to
the the small number of LBGs contained in the fields.
To show the differential contribution of LBGs, for a
3.2
In last sub-section we investigate the cross-correlation coefficient by specifying a smoothing scale = 10 . In this
sub-section we investigate the variety of the correlation coefficient at different angular scales. We re-define the correlation coefficient in frequency domain via the power spectrum
Rq () = p
PIRG ()
,
PIR () PG ()
(5)
Figure 3. Correlation coefficient between the 3.6 m sourcesubtracted NIRB and LBG flux maps vs. H-band limiting magnitude for three different map areas. Filled regions are the 1 ranges (68.3% probability). All fields are smoothed on scale
= 10 . For each signal map we run 30 contamination realizations: hence, for example, the (1.2)2 deg2 case uses 120 realizations.
The color
,
F3.6
h(F3.6 F1.6 ) i
Figure 4. Contribution to correlation coefficient from LBGs
with > z. The 1- variance is plotted by error bars. For displaying
purpose we slightly shift the x-positions.
the shot noise which dominates the small scale and the clustering term which dominates the large scale. The shot noise
term is from the same sources that contribute to both the
source-subtracted NIRB and the LBG flux map. This term
<
is dominant on small scales
100 . On larger scales, the
clustering term progressively takes over. The clustering term
arises from all sources sharing the same large scale struc-
(6)
(7)
Yue et al.
deconvolve the PSF from each map. We skip this step here
and directly calculate the magnitude difference on the map
without PSF convolution. Again we specify = 10 . The
predicted color as a function of Hlim is reported in Fig. 6,
allowing us to conclude that the magnitude difference, of
order of -0.13 mag, could be detected by cross-correlating
>
surveys with area
(0.3)2 deg2 . The figure reiterates that
smaller area fields would be affected by bias effects.
the contamination: it is the field-to-field variance of the correlation itself. We do not model errors introduced by observations, for example the mask effects. However, at least
theoretically we have shown that the contribution from the
faintest galaxies could be isolated from the NIRB through
the cross-correlation analysis. We pointed out that it is still
challenging to use the cross-correlation arising from clustering term in the cross-power spectrum to study galaxies
unresolved not only in NIRB observations, but also in LBG
>
surveys. This term is dominant at
200 , however even if
the survey area is as large as (2.4)2 deg2 , the signal still has
<
small significance, with a S/N ratio
3.
Additionally, our result have interesting implications for
the color of high-z galaxy populations. For LBGs are detected at wavelengths < 1.6 m while too faint to be resolved
for existing telescopes at 3.6 and 4.5 m, their mean color
in those band can be obtained by using the cross-correlation
with the NIRB.
It is worth noting that the predictions presented in this
paper assume a contamination in the form of NIRB fluctuation excess which could originate from an exotic population
of sources at even higher-z (Yue et al. 2013b). If however the
excess is found to arise more locally, we might gain the ability to model or subtract it more accurately, thereby making
the signal calculated in this paper more easily detectable,
>
hRi
0.1.
Importantly, this type of NIRB study can ultimately be
pushed further without requiring prior LBG detections, in
order to infer properties of still fainter galaxy populations
at higher-z that are inaccessible to direct detections with
any current instrument. This could, for example, be accomplished in a Lyman-break tomography study designed to
isolate high-z populations via multi-band cross-correlations
(Kashlinsky et al. 2015).
ACKNOWLEDGMENTS
4
REFERENCES
Atrio-Barandela F., Kashlinsky A., 2014, ApJ, 797, L26
Blaizot J., Wadadekar Y., Guiderdoni B., Colombi S. T., Bertin
E., Bouchet F. R., Devriendt J. E. G., Hatton S., 2005, MNRAS, 360, 159
Bouwens R. J. et al., 2011, ApJ, 737, 90
Bouwens R. J. et al., 2014, ApJ, 793, 115
Bouwens R. J. et al., 2015, ApJ, 803, 34
Cappelluti N. et al., 2013, ApJ, 769, 68
Choudhury T. R., Ferrara A., 2007, MNRAS, 380, L6
Cooray A., Bock J. J., Keatin B., Lange A. E., Matsumoto T.,
2004, ApJ, 606, 611
Cooray A., Gong Y., Smidt J., Santos M. G., 2012a, ApJ, 756, 92
Cooray A. et al., 2012b, Nature, 490, 514
Cooray A. et al., 2007, ApJ, 659, L91
Cooray A., Yoshida N., 2004, MNRAS, 351, L71
Fernandez E. R., Dole H., Iliev I. T., 2013, ApJ, 764, 56
Fernandez E. R., Iliev I. T., Komatsu E., Shapiro P. R., 2012,
ApJ, 750, 20
Fernandez E. R., Komatsu E., 2006, ApJ, 646, 703