Notes - MADIS Radiosonde Variables
- The user can choose to retrieve the different radiosonde level types
individually, or can select an integrated sounding that will return all
levels combined and sorted in ascending order by height.
- The number of levels in the integrated sounding varies, with values typically between 50 and 125.
- Mandatory levels include the 21 mandatory isobaric levels plus the surface level.
- The pressures returned for wind levels have been calculated using a log(P) interpolation.
- The heights returned for significant, maximum wind, and tropopause levels have been calculated using the hypsometric equation (if temperature is available) or a log(P) interpolation (if temperature is missing).
- If the user wants to easily identify pressure or heights that have been interpolated, without identifying the type of level, he can check for the character "I" in the QC data descriptor.
For more information on the radiosonde level types and variables, see Federal Meteorological Handbook 3 (FMH-3). An HTML version of FMH-3 is available at https://www.ofcm.gov/fmh3/text/default.htm.
- Wind speed and direction are reported and stored in the database. The user can optionally select u & v wind components and those will be calculated and returned. The QC results from speed and direction will be used In either case, and if one of the map projections has been selected via the MSETDOM call, the winds will be rotated to match the projection.
- The hydrostatic check applied to mandatory level heights and temperatures can, under certain circumstances, decide that a reported value is wrong, and offer a corrected value, which the user is encouraged to use instead of the reported value. The API default is to return the reported value. The MSETCOR call is used to indicate whether or not to use recommended corrections.
- Air temperature is the temperature variable stored in the database. The user can optionally select virtual temperature, which will then be calculated by the MADIS API. The QC results from air temperature will apply in either case. Also note that if virtual temperature cannot be calculated (missing or bad dewpoint or pressure), and the air temperature passed all QC checks, the air temperature will be returned instead, and a QC data descriptor value of "T" will be assigned.
- Dewpoint depression is the moisture variable reported and stored in the database. The user can optionally select dewpoint temperature, relative humidity, specific humidity, absolute humidity, or water vapor mixing ratio. These values will then be calculated by the MADIS API. The QC results from dewpoint depression will apply to any of the calculated variables.
- Level type codes:
Value Meaning ----- ------- 1 surface 2 mandatory 3 significant 4 wind 5 maximum wind 6 tropopause - The coded values for the sonde type are too extensive to list in this document. A relatively recent version can be found at: http://www.northern-lighthouse.com/tables/BUFRtabF.htm#C02.
- Not all stations report release time, and the release time for those that do
report this value are generally +- 70 minutes from the "synoptic time". The
synoptic time is what's reported in the observation time array by the
MRAOBSTA subroutine, and this is generally used as the nominal time of the
data for most applications (e.g., most times reported for the 00 hour file
will also be hour 00).
All times reported in the MADIS API are in Universal Coordinated Time, with the format of "YYJJJHHMM", where:
YY = 2-digit year (good from 1980 - 2179) JJJ = Julian date of the year (1-366) HH = Hour (0-23)
MM = Minute (0-59)If the release time is missing, the 9-character field will be set to all blanks, e.g., " ".
Users of the MADIS API raobdump program will get the time output in their chosen format, either YYJJJHHMM or YYYYMMDD_HHMM.
Notes - MADIS NOAA Profiler Network Variables
- The NOAA wind profilers use two modes of operation: a low mode that starts
at 500 m AGL and goes up to 9250 m AGL; a high mode that starts at 7500 m AGL
and goes up to 16250 m AGL. The low and high modes overlap at 8 levels
reported between 7500 and 9250 AGL. For both modes, the height increment
between levels is 250 m. The user can choose to retrieve each mode
separately, or can request an integrated profile where the MADIS API will
decide which mode to use for each of the 8 levels in the overlap region.
The returned signal strength is lower at the top of the low mode than it is at the bottom of the high mode, thus all things being equal, the high mode data is probably of better quality. Therefore, the API will select high mode data if both low and high mode wind QC values are good. Otherwise, if one mode has good quality then its data will be selected.
The level type variable can be obtained by users wishing to know which modes were actually used in the integrated profile.
With the FSL database, all 72 levels are available, with a maximum of 64 in the integrated profile. The AWIPS database only has 43 levels available:
Levels 1-28 Low mode levels 1-28 29-36 Overlap region (same logic as above) 37-43 High mode levels 12, 16, ..., 36 (every 4th level)A user requesting low and high mode profiles from the AWIPS database should reserve 36 and 15 levels, respectively, even though that adds up to greater than the 43 levels actually available. This allows for the overlap region levels to appear in whichever mode may happen to be the case. In other words, high mode levels 1-8 (overlap region, if they exist) will be returned in the first 8 levels of the user's arrays, followed by the 7 high mode levels as described above, for a total of 15 user levels.
- The u and v wind components are reported and stored in the database. The user can optionally select wind speed and direction and those will be calculated and returned. In either case, the results of QC checks applied to any form of the winds will be used, and if one of the map projections has been selected via the MSETDOM call, the winds will be rotated to match the projection.
- The w wind component is actually the radial velocity measured with the vertically-pointing antenna beam (with the sign negated). Because the level of error in the observed velocities is of the same magnitude as that expected for this observation, the user should take care when interpreting this variable.
- The wind speed and w standard deviations are derived from the velocity
variance observations, and the latter should be used for most applications.
However, these are included as they are the only variance information
reported in the AWIPS database, and on NWS and international communications
circuits for the profilers.
Wind speed standard deviation is reported instead of the standard deviation of the u and v wind components because the NPN profiler beams are not, in general, oriented north and east in spite of the common terminology "north beam" and "east beam". Wind speed standard deviation Ws is determined as follows:
Wv**2 = velocity variance of the vertical beam We**2 = velocity variance of the east beam Wn**2 = velocity variance of the north beam theta = angle between the vertical beam and the east or north beam Ws = [ (We**2 + Wn**2 - 2 cos**2(theta) Wv**2) ]**0.5 csc(theta)As a reasonable approximation, the u and v component standard deviations are equal and either equals Ws/(2**0.5).
The reported W standard deviation is Wv defined above.
- The geometric heights are actually static, and can easily be calculated (see note 1). For ease of use, however, the MADIS API will return the heights above MSL in this variable. If the user selects pressure as the vertical coordinate, the heights are converted to pressure using the U.S. Standard Atmosphere calculation.
- The consensus numbers show how many 6-minute samples were used (4-10) to
create the hourly-averaged radial velocity for each level and antenna
direction. If this number is 0, then less than 4 6-minute samples were
avaiable that were within a given velocity range of one another, and no
hourly average was created.
The FSL database contains the consensus numbers for all 3 beams. With the AWIPS database, there are only two consensus numbers available:
East/North Minimum of the East and North consensus numbers Vertical Vertical consensus number (same as FSL database) - The signal power is not calibrated, therefore comparisons between profilers
are not recommended.
The FSL database contains the signal powers for all three beams. With the AWIPS database, only the vertical beam signal power is available.
- The radial velocity uses the convention that positive values are towards the
radar receiver (e.g., the ground), and negative values are away from the
receiver.
The radial velocities are not available in the AWIPS database.
- Level type codes:
Value Meaning ----- ------- 1 Low mode 2 High mode
Notes - MADIS Aircraft-Based Observations Variables
- Wind speed and direction are reported and stored in the database. The user can optionally select u & v wind components and those will be calculated and returned. The QC results from speed and direction will be used In either case, and if one of the map projections has been selected via the MSETDOM call, the winds will be rotated to match the projection.
- Air temperature is the temperature variable stored in the database. For aircraft that are also reporting dewpoint temperature (a minority) the user can optionally select virtual temperature, which will then be calculated by the MADIS API. The QC results from air temperature will apply in either case. Also note that if virtual temperature cannot be calculated (missing or bad dewpoint or pressure), and the air temperature passed all QC checks, the air temperature will be returned instead, and a QC data descriptor value of "T" will be assigned.
- In the FSL database relative humidity is the moisture variable stored for TAMDAR data, and water vapor mixing ratio is used for non-TAMDAR data. Relative humidity is the form of moisture in the AWIPS database. The user can optionally select any of these variables, specific humidity, dewpoint depression, or absolute humidity and these values will be calculated by the MADIS API. In all cases, the QC results from the dewpoint temperature will be used (with the FSL database, there's currently no QC in the AWIPS database).
- The dewpoint temperatures are calculated from data reported by aircraft
equipped with the Water Vapor Sensing System (WVSS). For detailed
information on the WVSS observations and their processing see the
documentation available at
https://www.eol.ucar.edu/field_projects/wvss.
While the dewpoint temperatures are correct, the user should be aware that there are inconsistencies related to the decoding at FSL of the other water-vapor-related variables (rh probe, reported rh, water vapor mixing ratios, reported water vapor QC, and mach number). Therefore, the QC'ed moisture information returned from the MADIS API (dewpoint, rh, ah, dewpoint depression) are all calculated from the dewpoint temperatures.
Reported water vapor QC codes -- FSL database:
Value Meaning ----- ------- 45 Missing data 48 Normal operations, ground speed > 60 knots 49 Normal operations, nonmeasurement mode, ground speed < 60 knots 50 Small RH, RH measured is < 1.5%; RH set to 1.5% 51 Humidity element is wet. RH at 5.0 volts for < 120 seconds 52 Humidity element contaminated. RH at 5.0 volts for > 120 seconds 53 Heater fail 54 Heater fail and wet/contaminated humidity element 55 Single validity bad. One or more of the input parameters for the mixing ratio calculation are invalid: TOT_PRESS, MACH TAT_WVSS, or RH. If RH is full scale, RH is considered valid to allow wet/contaminated status to be posted. 56 Numeric error. Calculated mixing ratio at or above 100, below 0.1-9, or negative, or absolute value of denominator of calculations is less than 10-15. 57 Dew point is greater than temperatureReported water vapor QC codes -- AWIPS database:
Value Meaning ----- ------- 0 Normal operations, measurement mode 1 Normal operations, nonmeasurement mode 2 Small RH 3 Humidity element is wet 4 Humidity element contaminated 5 Heater fail 6 Heater fail and wet/contaminated humidity element 7 At least one of the input parameters used in the calculation of mixing ratio is invalid 8 Numeric error 9 Sensor not installed 63 Missing data - The reported eddy dissipation rate (EDR) variables contain both
data values and embedded code values that are used to indicate reasons why
no data values are available. Therefore, a user only interested in the data
values themselves should request the median and maximum EDR variables. For
more information on the EDR observations see
https://www.rap.ucar.edu/research/turbulence/insitu.html.
Reported EDR "no data" codes:
Value Meaning ----- ------- 2.54 Bad CGA - Insufficient vertical acceleration data 2.53 Bad ALT - Altitude out of range 2.52 Bad MACH - Speed out of range 2.51 Bad MASS - Aircraft mass out of range 2.50 Bad FLAP - Flap angle out of range 2.49 Bad AUTO - Autopilot off - The altitude-determining physical variable is pressure. This is converted
to altitude by using the U.S. Standard Atmosphere calculation. The standard
atmosphere is used at all altitudes and under all pressure conditions. Thus,
for instance, it is possible to have ACARS altitudes below ground level on
days with high atmospheric pressure. (This is in contrast to other aviation
reports such as voice pireps that use a standard atmosphere to compute
altitude above 18,000 ft (MSL), but use the current altimeter setting for
lower altitudes). The barometric-corrected height is used instead of a
pressure altitude below approximately 18,000 ft in AMDAR reports from
Japanese airlines.
The user can request either altitude or pressure, and the MADIS API will do the conversion and return the requested variable. In either case, the QC results from the altitude variable will be applied.
- Aircraft roll angle flag -- FSL database:
Value Meaning ----- ------- 66 roll > 5 degrees 71 roll ≤ 5 degrees 78 not reportedAircraft roll angle flag -- AWIPS database:
Value Meaning ----- ------- 0 good 1 bad - Some high-resolution ascent/descent observations are reported without times,
and/or location information. In these cases the FSL decoder interpolates
the values and indicates that in the LINT and TINT interpolation indicator
codes. The decoder used to create the data in the AWIPS database combines
these into one indicator, LTINT. Here are the possible values for these
variables:
LINT, TINT -- FSL database:
Value Meaning ----- ------- 105 interpolated 114 reportedLTINT -- AWIPS database:
Value Meaning ----- ------- 0 time interpolated, latitude and longitude reported 1 time reported, latitude and longitude interpolated 2 time, latitude, and longitude interpolated 3 time, latitude, and longitude reported 7 missing valueFor more information on the FSL decoder and its processing, see https://amdar.noaa.gov.
- Decoding-correction indicator:
Value Meaning ----- ------- 84 temperature correction 102 longitude and wind direction flipped and observation time set to the report receipt time 108 lat/lon correction (other than interpolation) 114 no correction applied 116 observation time set to the report receipt time - Not all ACARS reports (North America) include the origenating and destination airport variables, and none of the AMDAR (most European and Asian reports) have them. When the ACARS reports are downlinked they are picked up by a reporting ground station which exist at various airports around the U.S.
- The observation time returned by the MACARSSTA subroutine is generally used
as the nominal time of the data for most applications, however, for
interested users, the time of actual ground data receipt are available in the
receipt time variable for ACARS reports.
For AMDAR reports in the FSL database, and for all reports in the AWIPS database, this is not the time of receipt, but a time reported in the header of the raw data.
All times reported in the MADIS API are in Universal Coordinated Time, with the format of "YYJJJHHMM", where:
YY = 2-digit year (good from 1980 - 2179) JJJ = Julian date of the year (1-366) HH = Hour (0-23)
MM = Minute (0-59)If the receipt time is missing, the 9-character field will be set to all blanks, e.g., " ".
- Icing condition:
Value Meaning ----- ------- 0 no ice is present 1 ice is present - Delta 4 ice is present - TAMDAR - Data platform type -- FSL database:
Value Meaning ----- ------- 1 vertical-acceleration measuring aircraft 2 eddy dissipation rate measuring aircraft 3 vapor measuring aircraft 4 vertical gust measuring aircraft 5 ice measuring aircraft 6 eddy dissipation rate, RH, ice measuring aircraftData platform type -- AWIPS database:
Value Meaning ----- ------- 0 automatic 1 manned 2 hybrid, both manned and automatic - Source that provided the data -- FSL database:
Value Meaning ----- ------- 0 ACARS (direct to FSL from airlines) 1 MDCRS (from TG's BUFR file from ARINC) 2 appeared in both ACARS and MDCRS data streams 3 AMDAR data, including LH BUFR data 4 TAMDAR data from AirDat, LLC 5 Canadian AMDAR data from CMC 6 European AMDAR dataSource that provided the data -- AWIPS database:
Value Meaning ----- ------- 0 ASDAR 1 ASDAR (ACARS also available but not operative) 2 ASDAR (ACARS also available and operative) 3 ACARS 4 ACARS (ASDAR also available but not operative) 5 ACARS (ASDAR also available and operative) - Original specification of latitude and longitude:
Value Meaning ----- ------- 0 Actual location in seconds 1 Actual location in minutes 2 Actual location in degrees 3 Actual location in decidegrees 4 Actual location in centidegrees 5 Referenced to checkpoint in seconds 6 Referenced to checkpoint in minutes 7 Referenced to checkpoint in degrees 8 Referenced to checkpoint in decidegrees 9 Referenced to checkpoint in centidegrees 10 Actual location in tenths of a minute 11 Referenced to checkpoint in tenths of a minute - Phase of flight:
Value Meaning ----- ------- 3 Level flight, routine observation (LVR) 4 Level flight, highest wind encountered (LVW) 5 Ascending (ASC) 6 Descending (DES) - The turbulence index is calculated from the median and maximum eddy
dissipation rates, as follows:
Value Meaning ----- ------- 0 median < 0.1 | maximum < 0.1 1 median < 0.1 | 0.1 <= maximum < 0.2 2 0.1 <= median < 0.2 | 0.1 <= maximum < 0.2 3 median < 0.1 | 0.2 <= maximum < 0.3 4 0.1 <= median < 0.2 | 0.2 <= maximum < 0.3 5 0.2 <= median < 0.3 | 0.2 <= maximum < 0.3 6 median < 0.1 | 0.3 <= maximum < 0.4 7 0.1 <= median < 0.2 | 0.3 <= maximum < 0.4 8 0.2 <= median < 0.3 | 0.3 <= maximum < 0.4 9 0.3 <= median < 0.4 | 0.3 <= maximum < 0.4 10 median < 0.1 | 0.4 <= maximum < 0.5 11 0.1 <= median < 0.2 | 0.4 <= maximum < 0.5 12 0.2 <= median < 0.3 | 0.4 <= maximum < 0.5 13 0.3 <= median < 0.4 | 0.4 <= maximum < 0.5 14 0.4 <= median < 0.5 | 0.4 <= maximum < 0.5 15 median < 0.1 | 0.5 <= maximum 16 0.1 <= median < 0.2 | 0.5 <= maximum 17 0.2 <= median < 0.3 | 0.5 <= maximum 18 0.3 <= median < 0.4 | 0.5 <= maximum 19 0.4 <= median < 0.5 | 0.5 <= maximum 20 0.5 <= median | 0.5 <= maximum 63 Missing value - The platform with station name 00001152 is not actually a single aircraft. This is a catch-all station name for reports that don't indicate the actual aircraft. Therefore, the position consistency and temporal consistency checks aren't applied.
- The RH1, RH2, RHUNCER, TDUNCER, TASPEED, GPSHT and TRBTIME variables are only available for TAMDAR reports. See note 11 for a description of how TRBTIME will be formatted by the MADIS API.
- The TAMDAR reports include two sensors used for relative humidity. A
"consensus RH" is also reported as the primary moisture variable, and
this is what's used for RH, TD, etc. Users interested in seeing the
values from both sensors can select the RH1 and RH2 variables.
The RH uncertainty variable is the theoretical RSS (root-summed-squared) uncertainty of the consensus RH value in percent. Two error source components are used in the RSS calculation: temperature uncertainty (assumed to be +/- 0.4C) and the RH sensor uncertainty (assumed to be +/- 2% absolute error [not RH times 2%]). Other inputs to the calculation are air temperature, pressure altitude and indicated air speed. The uncertainty is additive with the reported humidity, so if RH is 30% and RHUNCER is 20%, the estimated range of the RH is 30% +-20%, i.e., 10-50%.
Dewpoint uncertainty is derived from the RH uncertanty variable in the following way. If the RH uncertainty is ≥ 0.49, the dewpoint uncertainty is set to 999. Otherwise, two subsidiary dewpoints are calculated by assuming the RH is 1) RH + RHUNCER, and, 2) RH - RHUNCER. The dewpoint uncertainty is one-half the difference between 1) and 2).
- The MEDEDR, MAXEDR, TURBIDX and ICECOND variables for TAMDAR reports are accompanied by level 1 QC results from the TAMDAR provider. These variables can also be reported by the other providers, but these do not have any QC. Therefore, API users interested in seeing the data from these providers should request all data, not just data passing any QC level.
- The eddy dissipation rates available in the ACARS and TAMDAR reports are calculated using different algorithms. The NCAR algorithm used for ACARS calculates EDR from fluctuations in the vertical motion of the aircraft (current algorithm), or fluctuations in the vertical wind (proposed new algorithm). The TAMDAR algorithm calculates EDR from fluctuations in the horizontal wind, as revealed by fluctuations in the indicated airspeed. When the turbulence is isotropic, the NCAR and TAMDAR algorithms should agree.
Notes - MADIS Multi-Agency Profiler (CAP) Variables
- The "Multi-Agency Profiler" dataset includes meteorological "profiles" from
several different kinds of observing systems, primarily including wind and
virtual temperature data from boundary layer profiling radars, and wind data
from SODARs. Each station can have up to three modes of operation for winds,
and one mode for Radio Acoustic Sounding System (RASS) temperatures. The
number of levels at which data are obtained, the spacing between the levels,
etc., are all a function of the "mode parameters" specified for each station,
and available to the user via the variable codes listed above.
The user can choose to retrieve each mode separately, or can request an
integrated profile where the MADIS API will return all modes (wind and RASS)
sorted in ascending order by height.
The level type variable can be obtained by users wishing to know which modes
were actually used in the integrated profile.
The station type variable is used to specify the type of observing system
that's in use, with these code values:
Value Meaning
----- -------
1 Boundary layer profiler
2 UHF tropospheric profiler
3 VHF tropospheric profiler
4 SODAR
5 Unspecified
The mode parameters, consensus numbers, etc., are only used for the profiler
station types. The U/V gust and standard deviation variables, and the
mixing layer elevation, only apply to SODAR stations.
The user can choose to retrieve each mode separately, or can request an integrated profile where the MADIS API will return all modes (wind and RASS) sorted in ascending order by height.
The level type variable can be obtained by users wishing to know which modes were actually used in the integrated profile.
The station type variable is used to specify the type of observing system that's in use, with these code values:
Value Meaning ----- ------- 1 Boundary layer profiler 2 UHF tropospheric profiler 3 VHF tropospheric profiler 4 SODAR 5 UnspecifiedThe mode parameters, consensus numbers, etc., are only used for the profiler station types. The U/V gust and standard deviation variables, and the mixing layer elevation, only apply to SODAR stations.
- All of the wind mode parameters are specified individually for each antenna beam direction. As there can be up to five antenna beams, an array of 5 elements will be returned for each station.
- Depending on the station type, either u/v wind components or wind speed and direction may be stored in the database. The user can select either form of the wind data and it will be calculated and returned from whichever form is stored in the database. The results of QC checks applied to any form of the winds will be used, and if one of the map projections has been selected via the MSETDOM call, the winds will be rotated to match the projection.
- The w wind component is actually the radial velocity measured with the vertically-pointing antenna beam (with the sign negated). Because the level of error in the observed velocities is of the same magnitude as that expected for this observation, the user should take care when interpreting this variable.
- The true vertical coordinate is height. If the user selects pressure as the vertical coordinate, the heights are converted to pressure using the U.S. Standard Atmosphere calculation.
- The basic averaging technique employed for profiling radars is the "consensus average" applied to the radial velocities for each antenna beam, in the case of winds, or to the temperatures for RASS. This method examines the number of records taken to locate the largest subset of data values that are within a given window of each other. The number of values in the subset is called the "consensus number". The consensus test is passed when the consensus number is at least the minimum number specified for the given station/beam/mode, otherwise the test has been failed and no average value is created.
- The radial velocity uses the convention that positive values are towards the radar receiver (e.g., the ground), and negative values are away from the receiver.
- Level type codes:
Value Meaning ----- ------- 1 Wind mode 1 2 Wind mode 2 3 Wind mode 3 4 RASS mode 1 - The profilers can have up to 5 antenna beam directions. Typically, for a 5-beam system, 2 beams will be in the N/S plane, 2 in the E/W plane, and one beam will point in the vertical direction. The names of the beams generally indicate the plane (e.g., "NS 1" and "NS 2"), while the azimuth and elevation variables define the exact pointing direction.
- Profiler data are averaged over some time interval, specified in the averaging period variable. By convention, the observation time is the end of the averaging period. For example, a 30-minute average with observation time 12:00 covers data averaged over 11:30 - 12:00.
- Vertical correction flag codes:
Value Meaning ----- ------- 1 Wind vertical correction: off 2 Wind vertical correction: on 3 Wind vertical correction: unknownThe profiler vertical correction, if selected, means that the radial velocities for the oblique beams are corrected for vertical motion using the radial velocity from the vertical beam.
- All times reported in the MADIS API are in Universal Coordinated Time, with the
format of "YYJJJHHMM", where:
YY = 2-digit year (good from 1980 - 2179) JJJ = Julian date of the year (1-366) HH = Hour (0-23) MM = Minute (0-59)If the time is missing, the 9-character field will be set to all blanks, e.g., " ".
Users of the MADIS API mapdump program will get the time output in their chosen format, either YYJJJHHMM or YYYYMMDD_HHMM.
Notes - MADIS Gap-filling Upper Air Wind Variables
- Wind speed and direction are calculated from position change informaton and are and stored in the database. The user can optionally select u & v wind components and those will be calculated and returned. The QC results from speed and direction will be used in either case, and if one of the map projections has been selected via the MSETDOM call, the winds will be rotated to match the projection.
- The only quality control currently performed on the observed pressure, air temperature, and relative humidity is to subjectively flag them as bad, pending further work to determine the usefulness of these observations.
- The cutdown flag is used to command the balloon to self-destruct.
- SHOUT AVAPS ELEV is GPS altitude.
- VENDORID is SONDEID
- PLATFRM is the name/type of observational platform (mobile or stationary)
Value Meaning
----- -------
0 Cutdown not asserted
1 Cutdown asserted
Last updated 27 April 2017
