GS - EXOSAT GSPC Spectra and Lightcurves

Overview

The results and data products from the EXOSAT GSPC. Only sources with ME count rates of at least 5 ct/s/half are included.

References

Peacock A. et al., 1981, Sp. Sci. Rev., Vol. 30, p. 525.

The gas scintillation proportional counter (GSPC) has a dE/E of 4.5(E/6 keV)**-0.5 percent FWHM, approximately a factor of 2 better than obtained by the ME experiment. The peak effective area is 100 cm**2, a factor of 8 less than one ME array half. The field of view is the same as the ME. The energy spectra were pulse height analyzed into 256 channels covering either the 2-32 keV or 2-16 keV energy bands, depending on the gain. The gain was accurately monitored using two background line features at 10.54 and 12.70 keV caused by fluorescence in the lead collimator.

Data Products

All the products available from this database, spectra, response matrix and lightcurves has been converted at HEASARC in FITS format.

A plot of the lightcurves and spectra is provided also in GIF format. The available products are :

 * a background subtracted average spectra and the associated response
   matrix for each observation (one file in FITS and GIF format).

 * two lightcurves, background subtracted and deadtime corrected, with a
   typical time resultion of 8 seconds in the two energy bands: 2-8 keV
   and 8-15 keV (two files in FITS format and one in GIF format
   showing the lightcurves for both energy bands).

All type I bursts are excluded from the source spectra, but are still kept in the light curves.

Automatic Analysis

The GS auto was run for all observations where the ME count rate was in excess of 5 ct/s/half. Sources with count rates lower than this do not have sufficient signal-to-noise to justify a GS analysis. The GS products were generated by an automatic processing described below:

 * Creation of a light curve in the energy range 2-8 keV and 8-15 keV with a
   time resolution of 8 s (or less when the original time resolution was
   <8 s) for the slew on or off source.

 * Creation of a spectrum for the slew on or off source, excluding sources
   seen in the slew light curves.

 * Creation of a light curve in the energy range 2-8 keV and 8-15 keV with a
   time resolution of 8 s (or less when the original time resolution was
   < 8 s) for the source.

 * Creation of a spectrum for entire observation of the source (excluding
   any bursts that occurred).

 * Subtraction of the background.  For faint sources two lead line features
   in the source and background spectrum are used to calculate the
   gain conversion factor. In bright sources where these lines are hard to
   detect, the gain conversion factor is determined using the lead line
   features in the spectrum from the slew on or off source. The standard
   background is then adjusted corresponding to the gain conversion factor,
   re-normalized to the slew spectrum and subtracted.

 * When the 3-10 keV count rate is >0.5 counts/s the spectrum is analyzed
   using XSPEC.

Spectral Parameters

The best fit `spectral model` parameter gives the number of the model that has the smallest reduced chi**2, written to the `chisqr` parameter. The `spectral model` number can be between 0 and 8 representing the following models:

 * 0 - no spectral fit made
 * 1 - thermal bremsstrahlung
 * 2 - power law
 * 3 - blackbody
 * 4 - compton E**-gamma * exp(-E/kT).

The best fitting parameters for models 1 -- 4 are each given with and without an iron K line present. The parameters with a line present are indicated by `PL_G chisqr`, `PL_G norm` etc. (`_G` indicates the inclusion of a gaussian line). The more complex models are:

 * 5 - thermal bremss. plus line (TH_G)
 * 6 - powerlaw plus line (PL_G)
 * 7 - blackbody plus line (BB_G)
 * 8 - compton plus line (CO_G).

In some cases, especially for faint sources, more than one model may give an acceptable fit.

The best fit parameters for each model and the corresponding reduced chi**2 are written to the GS database. These parameters are:

 PL chisqr   PL norm   PL gamma   PL Nh
 TH chisqr   TH norm   TH kT      PL Nh
 BB chisqr   BB norm   BB kT      BB Nh
 CO chisqr   CO norm   CO gamma   CO kT    CO Nh

The presence of an iron K line in the best fit model is indicated when the parameter `line present` is set to 1, rather than 0. In this case the equivalent width (in eV) and energy of the line (in keV) are to be found in `EW line` and `energy line`, respectively. The line width for the GS are given in terms of one sigma, rather than FWHM as in the ME. For models 5 -- 8 the iron line parameters are written to the parameters `TH_G E line`, `TH_G Sigma line` etc.

The spectral parameters for the best-fit model are stored in the database as `P1`, `P2`, etc. The contents of each of these parameters depend on the number of the best fit model as follows:

 Best fit model          P1      P2      P3      P4      P5      P6     P7
 --------------          --      --      --      --      --      --     --
 Thermal                 Nh      kT      norm
 Powerlaw                Nh      gamma   norm
 Blackbody               Nh      kT      norm
 Compton                 Nh      gamma   kT      norm
 Thermal & Line          Nh      kT      norm    Eline   sigma  norml
 Powerlaw & Line         Nh      gamma   norm    Eline   sigma  norml
 Blackbody & Line        Nh      kT      norm    Eline   sigma  norml
 Compton & Line          Nh      gamma   kT      norm    Eline  sigma  norml

where:

 * norm  : spectral normalization
 * kT    : temperature in keV
 * gamma : powerlaw photon index
 * Nh    : absorption in units of 10**22 atoms cm**-2
 * Eline : energy of the line in keV
 * sigma : line width in sigma fixed at a value of 0.5 keV
 * norml : total integrated number of counts in the line.

The best fitting parameters in the GSPC database may differ from those in the ME database. This is because GS spectra have been fitted using the improved spectral fitting routine XSPEC, rather than by the older, less reliable program used for the ME spectra.

Quality

A quality flag is used to indicate the quality of the background subtraction. This has the same goodness values as the ME, with QF = 5 very good, QF = 0 terrible. In general 3-5 is good, while 0-2 should not be used for spectral analysis.

The GS automatic spectral results may be unreliable because

 * there may be more than one model that gives an acceptable fit

 * the automatic spectral fitting program sometimes was trapped in a
   false minimum, causing apparently the wrong model to give the best fit

 * entries with `Qflag GS` <3 indicate problems with background subtraction,
   which may affect the spectral fit.

Parameters

BB_Chisqr
Chisq for BB

BB_G_Chisqr
Chisq for BB & Gaussian

BB_G_E_Line
Energy Line for BB & Gaussian

BB_G_Ewidth
Equiv. Width for BB & Gaussian

BB_G_kT
Temp for BB & Gaussian

BB_G_NH
Nh for BB & Gaussian

BB_G_Norm
Norm for BB & Gaussian

BB_G_Norm_Line
Norm Line for BB & Gaussian

BB_G_Sigma_Line
Sigma Line for BB & Gaussian

BB_kT
Temp for BB

BB_NH
Nh for BB

BB_Norm
Norm for BB

Beta_Angle
The angle between the source and the Sun for the observation. If it is less than 90 degrees then there may be solar flare contamination, arising from solar X-rays leaking through a small pin hole in the side of the detector. Observations at less than beta 90 were mostly confined to the first few weeks of the mission.

Bg_Lightcurve1
Bg Cts/sec E-band 1

Bg_Lightcurve1_Error
Bg Error/sec E-band 1

Bg_Lightcurve2
Bg Cts/sec E-band 2

Bg_Lightcurve2_Error
Bg Error/sec E-band 2

BII
bII of source (degrees)

Blw1
The lower burst length window limit.

Blw2
The upper burst length window limit.

Chisqr
Reduced Chisq

Co_Alpha
Index for CO

Co_Chisqr
Chisq for Co

Co_G_Alpha
Index for CO & Gaussian

Co_G_Chisqr
Chisq for CO & Gauusian

Co_G_E_Line
Energy Line for CO & Gaussian

Co_G_Ewidth
Equiv. Width for CO & Gaussian

Co_G_kT
Temp for CO & GAussian

Co_G_NH
Nh for CO & Gaussian

Co_G_Norm
Norm for CO & Gaussian

Co_G_Norm_Line
Norm LIne for CO & Gaussian

Co_G_Sigma_Line
Sigma Line for CO & Gaussian

Co_kT
Temp for CO

Co_NH
Nh for CO

Co_Norm
Norm for CO

Colleff
The collimator efficiency.

Count_Rate
The count rate in the 3-10 keV band, corrected for dead time. Other count rates are available, viz.:

   * `Rate 1` - The count rate in band 1, 2-3 keV
   * `Rate 2` - The count rate in band 2, 3-6 keV
   * `Rate 3` - The count rate in band 3, 6-10 keV
   * `Rate 4` - The count rate in band 4, 10-20 keV
   * `Rate total` - The count rate in the entire band
   * `BG lightcurve1` - Slew background count rate in 2-8 keV band
   * `BG lightcurve2` - Slew background count rate in 8-15 keV band.

Count_Rate_Error
Count Rate Error in 3-10keV Band

DB_Entry
Entry no. in database

Dec
The Declination of the target.

Dtc_Factor
The dead time correction factor.

Duration
Duration (seconds)

End_Time
The end time of the observation. The units and use of this parameter are the same as those for `time`.

Energy_Line
Line Energy

Ew1
The lower energy channel limit.

Ew2
The upper energy channel limit.

Ew_Line
EW of line

Exposure
The `exposure` is the total on source observation time in seconds. This includes all dead time effects, interruptions in coverage, etc.

File_Bgspectrum
Bg Spectrum Filename

File_Lcurve
The 'FILE_LCURVE' paramater contains the file name of the 'C' lightcurve. The 'C' lightcurve is in the energy band 2-8 keV. This is also used the name of the GIF file.

File_Spectrum
The 'FILE_SPECTRUM' paramater contains the file name of the associated spectrum.

Gain
The energy gain setting for the observation.

Gcf_Factor
The energy gain correction factor for the observation.

Hp_Entry_Number
HP Number

HR
The hardness ratio (6-10 keV / 3-6 keV).

HR_Error
Error in Hardness ratio

ID_Number
Source order no.

Last_Update
time of last record change (MJD)

Lbgspectra
max length of bg spectra name in chars

Lightcurve2_Name
The 'LIGHTCURVE_LCURVE' paramater contains the file name of the 'D' lightcurve The 'D' lightcurve is in the energy band 8-15 keV.

LII
lII of source (degrees)

Line_Present
1=line, 0=no line in fit

Llcurve
max length of l/c names in chars

Lspectra
max length of spectrum name in chars

Macro_Name
Macro Name

Name
The `name` parameter gives the target name. This name usually corresponds to the target name as specified by the original observer. If a source has been observed by different observers it may be entered under two or more names. It is always recommended that any search by `name` is always followed by a search on the found coordinates.

Nbgspectra
Number of bg spectra

NH_21cm
Galactic NH

Nlcurve
Number of serial l/cs

Nspectra
Number of spectra

Num_Spec
Number of parameters

OBC_Program
The ID number of the OBC program (32 = HEBL4, 1 = DIRECT).

P1
Best fit model parameter 1

P10
Best fit model parameter 10

P2
Best fit model parameter 2

P3
Best fit model parameter 3

P4
Best fit model parameter 4

P5
Best fit model parameter 5

P6
Best fit model parameter 6

P7
Best fit model parameter 7

P8
Best fit model parameter 8

P9
Best fit model parameter 9

PI_Code
PI Code

PL_Alpha
Index for PL

PL_Chisqr
Chisq for PL

PL_G_Alpha
Index for PL & Gaussian

PL_G_Chisqr
Chisq for PL & Gaussian

PL_G_E_Line
Energy Line for PL & Gaussian

PL_G_Ewidth
Equiv. Width for PL & Gaussian

PL_G_NH
Nh for PL & Gaussian

PL_G_Norm
Norm for PL & Gaussian

PL_G_Norm_Line
Norm Line for PL & Gaussian

PL_G_Sigma_Line
Sigma Line for PL & Gaussian

PL_NH
Nh for PL

PL_Norm
Norm for PL

Point_Dec
The declination of the EXOSAT pointing.

Point_RA
The RA of the EXOSAT pointing.

Process_Date
time of macro run

Proposal_Code
Proposal Code

Proposal_Type
Observation Type

Qflag_Gs
GS background subtraction quality flag; 3-5 are acceptable. Do not use entries with `Qflag` = 1 or 2.

RA
The Right Ascension of the target.

Rate_1
Counts/s in band 1, 2-3 keV

Rate_1_Error
Error/s in band 1, 2-3 keV

Rate_2
Counts/s in band 2, 3-6 keV

Rate_2_Error
Error/s in band 2, 3-6 keV

Rate_3
Counts/s in band 3, 6-10 keV

Rate_3_Error
Error/s in band 3, 6-10 keV

Rate_4
Counts/s in band 4, 10-20 keV

Rate_4_Error
Error/s in band 4, 10-20 keV

Rate_Total
Counts/s in total band

Rate_Total_Error
Error/s in total band

Roll_Angle
Roll Angle (deg *10)

Sequence_Number
Obs. sequence number

Slew_Lightcurve1_Name
Slew Lightcurve E-band 1 Filename

Slew_Lightcurve2_Name
Slew Lightcurve E-band 2 Filename

Slew_Spectrum_Name
Slew Spectrum Filename

Spectral_Model
Best Fit model code

SR
The softness ratio (2-3 keV / 3-6 keV).

SR_Error
Error in Softness ratio

Th_Chisqr
Chisq for TH

Th_G_Chisqr
Chisq for TH & Gaussian

Th_G_E_Line
Energy Line fot TH & Gaussian

Th_G_Ewidth
Equiv. Width for TH & Gaussian

Th_G_kT
Temp for TH & Gaussian

Th_G_NH
Nh for TH & Gaussian

Th_G_Norm
Norm for TH & Gaussian

Th_G_Norm_Line
Norm Line for TH & Gaussian

Th_G_Sigma_Line
Sigma Line for TH & Gaussian

Th_kT
Temp for TH

Th_NH
Nh for TH

Th_Norm
Norm for TH

Time
The `time` of the observation refers to the start time. This is listed as yr.day, where yr is the last two digits of the year and day is the day of the year, e.g., day 300 of 1984 is 84.300. All times are stored internally as a short history file key (SHF) which is the number of seconds before or after 1 January 1980 at 00:00 hrs UT. Times are accurate to the nearest second.

Xspec_Done
XSPEC Summary File exists?

Contact Person

Questions regarding the GS database table can be addressed to the HEASARC Help Desk.

Page Author: Browse Software Development Team
Last Modified: Monday, 16-Sep-2024 17:28:31 EDT