NOTICE:
This Legacy journal article was published in Volume 1, May 1992, and has not been
updated since publication. Please use the search facility above to find regularly-updated
information about this topic elsewhere on the HEASARC site.
|
SSS - The Einstein Solid State
Spectrometer Database at the HEASARC
Steve Drake, Keith Arnaud, and Nick White
HEASARC
Overview
This article describes how to use BROWSE to access the Einstein Observatory
Solid State Spectrometer (SSS) catalog of observations and data products that
is available in the HEASARC database. It also gives an example of using XSPEC
to model the SSS spectrum of one of the objects observed by the SSS. For a
description of the instrument and an improved ice model, see the accompanying
article on the SSS calibration by Christian, Swank, Szymkowiak and White.
SSS Data Products Available in the HEASARC Database
There are 632 distinct SSS observations in the database. Each observation
generally consists of a small number of segments that have been concatenated to
improve the signal-to-noise. If a target was observed with the SSS over an
interval containing more than one calendar (UT) day, the usual procedure was to
create a different observation for the data accumulated on that target for each
day. The spectra are not background-subtracted. Detector response matrices and
background spectra are made using the VIMAT program, or from within BROWSE
using the XSPEC command. There is a background-subtracted 81.92 second time
resolution light curve in the 1.0-4.0 keV band covering the time interval of
each spectrum.
In addition to the primary SSS data products, there are also some associated
data products in the SSS database, namely near-simultaneous data obtained by
the Einstein Monitor Proportional Counter (MPC). The MPC operated in the 1-15
keV band, with eight pulse height channels. It was a mechanically collimated
detector with a full width half maximum field of view of 45 arc minutes. There
is one MPC spectrum (accumulated over roughly the same time interval as the SSS
observation and at times when the satellite was pointing at the SSS target) to
match each SSS spectrum, with a small number (29/632) of exceptions where there
were no valid MPC data. The MPC spectra are background-subtracted. There is one
MPC detector response matrix for the entire mission, which can be extracted
using the VIMAT program. In addition, there are also background-subtracted
40.96 second time resolution MPC light curves in the 1.0-15.0 keV band for this
same time interval. Notice that, due to the differing instrumental observing
constraints and particle background thresholds, the SSS and MPC data products
are only quasi-simultaneous. The start and stop times referred to in the
parameters description refer specifically to the SSS observations.
Data Selection
Users can select SSS data products using a variety of keys, including target
name, spatial coordinates, time of observation, and class. Searches by name
(sn) are always risky, since most targets will be known under a variety of
aliases, and the user might find a null result because either there was no SSS
observation of the object in question or because he used a different name for
the object than is used in the database. We have tried to select the most
commonly used name for each object in the SSS database since the original
naming convention for the SSS targets was fairly arbitrary and not completely
self-consistent, but this of course is somewhat subjective. Note that the names
in the database are concatenated (i.e., embedded blanks have been removed,
e.g., Cyg X-1 for Cyg X-1).
Using coordinate searches (e.g., the sc command) is always the most reliable
search method. Since the field of view of the SSS was small (6' in diameter)
the radius of the search cone should be adjusted accordingly. Notice also that
the FOV of the MPC was 45' and thus in some cases the dominant X-ray source in
the MPC FOV may lie outside the SSS FOV.
Other database parameters can be searched on using the sp command:
(a) Time: This is the start time of the observation in format "Year.Day".
(b) Type: This is a classification made by Jean Swank and Damian Christian as
to the nature of the source. The following types have been defined:
'bes' X-ray binary with Be star primary, e.g., Gam Cas.
'blk' X-ray binary with black hole primary, e.g., Cyg X-1.
'cluster' Cluster of galaxies, e.g., A 2029.
'cvs' Cataclysmic variable, e.g., AM Her.
'gal' Galaxy or quasar, e.g., 3C273.
'lmxb' Low-mass X-ray binary, e.g., Cyg X-2.
'pulsar' Pulsar, e.g., Cen X-3.
'rscvn' RS CVn active binary system, e.g., AR Lac.
'snr' Supernova remnant, e.g., sn 1006.
'star' Miscellaneous varieties of galactic stars not belonging to the
other stellar categories, e.g., tau Sco.
'mispoint' Observations made at incorrect coordinates.
(c) Class: This is the HEASARC BROWSE Classification of the target.
(The BROWSE classification scheme is described in detail in the HEASARC
document 'Available Databases').
Example of accessing SSS data using BROWSE and XSPEC
Log on to NDADSA in the usual way (see the accompanying article on the HEASARC
On-line Service). User inputs are preceded by > or :, our comments by !!,
and everything else is output of BROWSE program. Due to changes in the models
in the XSPEC program, the actual numerical results given here may not be
precisely reproducible.
HEASARC > browse sss
!! Now try and see if our favorite active star AR Lac was observed by the
SSS.
SSS_TOTAL_DEC > sn arlac
!! Notice the name of AR Lac is concatenated. Searches by name are always
risky; it is usually safer to search in a cone (sc)
!! about a known position.
name seq time expos count ice delt ra dec
(target) (#) (yr/day) (s) rate -ice (1950) (1950)
1 ARLAC 1740 78/335 901 0.9 3.52 0.00 22 06 38.0 45 29 46
2 ARLAC 1740 78/337 819 0.1 3.62 0.00 22 06 38.0 45 29 46
3 ARLAC 1740 78/338 819 0.8 0.69 0.12 22 06 38.0 45 29 46
4 ARLAC 1740 79/147 8192 0.7 1.31 0.01 22 06 38.0 45 29 46
5 ARLAC 1740 79/146 9666 0.5 1.23 0.02 22 06 38.0 45 29 46
6 ARLAC 1740 78/339 737 0.4 2.38 0.00 22 06 38.0 45 29 46
!! BROWSE found 6 observations: let's pick the longest one to examine.
SSS_TOTAL_NAM 6> dall 5
!! This lists all available parameters for the specific observation.
NAME ARLAC
TIME 1979/146 12:52
STOP TIME 1979/146 18: 9
YEAR 1979
EXPOSURE TIME 9666
RA (1950) 22 06 38.0
DEC (1950) 45 29 46.0
LII 95.55381011962891
BII -8.299847602844238
TYPE rscvn
FILE NAME sarlace
COUNT RATE 0.4780000
COUNT RATE ERROR 8.0000004E-03
COMMENT 323 28 29 30 31 32 33
ICE START 1.232168
ICE STOP 1.254181
DELTA ICE 1.7865416E-02
FLUX 0.0000000E+00
MIN COUNT RATE 7.0000002E-03
MIN COUNT RATE ERROR 4.6000000E-02
MAX COUNT RATE 0.8050000
MAX COUNT RATE ERROR 0.1060000
TCHI2 1.760000
SEQUENCE NUMBER 1740
REFS
FILE SPECTRUM sarlace
FILE LCURVE sarlace
CLASS 1900
SPARE
ASSOCIATED FILE LIGHTCURVE arlace
ASSOCIATED FILE SPECTRUM arlace
SSS_TOTAL_NAM 6> cpd /te
!! This command sets the current plot device to be a Tektronix plot. Use your
own plot device here; cpd ? gives a list of the available devices.
SSS_TOTAL_NAM 6> pp/sp 5
!! First plot the spectrum of this observation.
Plotting to /te
File sarlace.pha in current directory
!! Now plot the sss and mpc light-curves of this observation.
SSS_TOTAL_NAM 6> pp/li/asli 5
Plotting to /te
Figure 1
!! Notice that the SSS observations (top light curve) consists of 4 separate
segments spread over some five hours. Also notice
!! that the associated MPC light curve (bottom light curve) is only
'quasi-simultaneous'; some time intervals there are truly
!! simultaneous data from both detectors, while at other times there may only
be good data from one of them (or neither!).
!!
!! If you want to know what other RS CVn stars have been observed, use the
BROWSE command "rscvn" (which is an alias
!! for sp class 1900 1909).
SSS_TOTAL_NAM 6> rscvn
!! or
SSS_TOTAL_NAM 6> sp
Enter indexed parameter name (or ?, or exit): class
Enter minimum numeric value: 1900
Enter maximum numeric value: 1909
34
1 HR1099 3620 79/223 6062 1.4 0.92 0.04 03 34 13.0 00 25 29
2 ARLAC 1740 78/338 819 0.8 0.69 0.12 22 06 38.0 45 29 46
3 RSCVN 1741 78/338 1474 0.2 1.05 0.07 13 08 18.0 36 12 00
4 RSCVN 1741 79/152 2785 0.2 0.69 0.23 13 08 18.0 36 12 00
5 ARLAC 1740 79/146 9666 0.5 1.23 0.02 22 06 38.0 45 29 46
6 HKLAC 1747 79/156 1884 0.2 1.22 0.00 22 02 57.0 46 59 27
7 ARLAC 1740 78/335 901 0.9 3.52 0.00 22 06 38.0 45 29 46
.
.
.
33 HR1099 3620 79/032 737 0.6 1.79 0.00 03 34 13.0 00 25 29
34 ALPAUR 1726 79/061 7618 0.6 1.78 0.01 05 12 59.0 45 56 58
34 entries retrieved
!! To do a detailed spectral analysis of one of these spectra, let's say the AR
Lac observation that is entry 5 above,
!! we now spawn XSPEC from BROWSE.
SSS_TOTAL_CLA 34 > xspec/int 5
xspec 13:02:55 16-JAN-92
XSPEC> his SARLACE.XHS
XSPEC> data SARLACE.PHA;1
Net count rate (cts/cm2/s) for file 1 3.8270E-03+/- 5.7325E-05( 83.6% total)
Net correction flux for file 1= 9.5879E-04
Command not found; type ? for a command listing
XSPEC> ignore bad
File 1 Ignored channels 1 to 1
File 1 Ignored channels 86 to 94
XSPEC> show
Log file : LOG.LOG
History file : USR:[DRAKE]SARLACE.XHS;1
Information for file 1
belonging to plot group 1, data group 1, det id = EINSTEIN SSS
Current data file: USR:[DRAKE]SARLACE.PHA;1
Background file :USR:[DRAKE]SARLACE.BCK;1
Correction file :USR:[DRAKE]SARLACE.COR;1 with norm 0.0000
Response file : USR:[DRAKE]SARLACE.RSP;1
Noticed channels 2 to 85
File observed count rate 3.7036E-03+/-5.59956E-05 cts/cm^2/s
0.6667 +/-1.00792E-02 cts/s
After correction of 0.0000E+00; Model predicted rate: 0.0000E+00
XSPEC> p
!! This will plot the pha spectrum.
XSPEC> model raymond raymond wabs
!! RS CVn spectra generally require at least 2 thermal components to be well
fit. In this example, we are going to fit the
!! observed spectrum using 2 Raymond and Smith thermal plasmas plus
interstellar absorption.
Input parameter value, delta, min, bot, top, and max values for ...
Fit parameter 1 of component 1 1 raymond kT(keV)
1.000 1.0000E-02 8.0800E-03 8.0800E-03 79.90 79.90
2.0
Fit parameter 2 of component 1 2 raymond Abundanc
1.000 -1.0000E-03 0.0000E+00 0.0000E+00 5.000 5.000
1
Fit parameter 3 of component 1 3 raymond Redshift
0.0000E+00 -1.0000E-03 0.0000E+00 0.0000E+00 2.000 2.000
0
Fit parameter 4 of component 1 0 raymond norm
1.000 1.0000E-03 0.0000E+00 0.0000E+00 1.0000E+05 1.0000E+06
1
Fit parameter 5 of component 2 1 raymond kT(keV)
1.000 1.0000E-02 8.0800E-03 8.0800E-03 79.90 79.90
0.5
Fit parameter 6 of component 2 2 raymond Abundanc
1.000 -1.0000E-03 0.0000E+00 0.0000E+00 5.000 5.000
1
Fit parameter 7 of component 2 3 raymond Redshift
0.0000E+00 -1.0000E-03 0.0000E+00 0.0000E+00 2.000 2.000
0
Fit parameter 8 of component 2 0 raymond norm
1.000 1.0000E-03 0.0000E+00 0.0000E+00 1.0000E+05 1.0000E+06
1
Fit parameter 9 of component 3 1 wabs nH 10^22
1.000 1.0000E-03 0.0000E+00 0.0000E+00 1.0000E+05 1.0000E+06
1.0e-3
Summary of model parameters
Model Fit Model Component Parameter Value Flag
par par comp index type
1 1 1 1 raymond kT(keV) 2.00000 *
2 2 1 2 raymond Abundanc 1.00000
3 3 1 3 raymond Redshift 0.000000E+00
4 4 1 0 raymond norm 1.00000 *
5 5 2 1 raymond kT(keV) 0.500000 *
6 6 2 2 raymond Abundanc 1.00000
7 7 2 3 raymond Redshift 0.000000E+00
8 8 2 0 raymond norm 1.00000 *
9 9 3 1 wabs nH 10^22 1.000000E-03*
5 variable fit parameters
Current statistic= 5.1256E+07 using 84 PHA bins.
!! Try and fit the AR Lac spectrum with the above model.
XSPEC> fit 100
RENORM: renorm = 8.5087E-03
Chi-squared Lvl Fit param # 1 2 3 4
5 6 7 8 9
Fit parameter 9 has pegged
352.17 -2 2.766 1.000 0.0000E+00 1.8242E-02
0.5588 1.000 0.0000E+00 4.2370E-03 0.0000E+00
285.31 -3 2.934 1.000 0.0000E+00 1.7926E-02
0.6706 1.000 0.0000E+00 5.2860E-03 0.0000E+00
259.40 -4 3.619 1.000 0.0000E+00 1.6394E-02
0.7070 1.000 0.0000E+00 6.7262E-03 0.0000E+00
258.66 -5 3.963 1.000 0.0000E+00 1.6186E-02
0.7082 1.000 0.0000E+00 6.8234E-03 0.0000E+00
258.54 -6 4.314 1.000 0.0000E+00 1.6080E-02
0.7117 1.000 0.0000E+00 6.8961E-03 0.0000E+00
256.69 -7 4.450 1.000 0.0000E+00 1.5883E-02
0.7220 1.000 0.0000E+00 7.0421E-03 0.0000E+00
252.22 -8 5.428 1.000 0.0000E+00 1.4463E-02
0.7819 1.000 0.0000E+00 8.3123E-03 0.0000E+00
250.67 -9 6.756 1.000 0.0000E+00 1.4649E-02
0.7752 1.000 0.0000E+00 8.4332E-03 0.0000E+00
250.65 -10 6.682 1.000 0.0000E+00 1.4670E-02
0.7737 1.000 0.0000E+00 8.3730E-03 0.0000E+00
250.65 -11 6.681 1.000 0.0000E+00 1.4670E-02
0.7737 1.000 0.0000E+00 8.3727E-03 0.0000E+00
Unpegged 1 parameters
250.65 1 6.681 1.000 0.0000E+00 1.4670E-02
0.7737 1.000 0.0000E+00 8.3726E-03 0.0000E+00
1 1 1 raymond kT(keV) 6.68079 +/- 1.4447
2 1 2 raymond Abundanc 1.00000 +/- 0.00000E+00
3 1 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
4 1 0 raymond norm 0.146701E-01 +/- 0.44039E-03
5 2 1 raymond kT(keV) 0.773705 +/- 0.96893E-02
6 2 2 raymond Abundanc 1.00000 +/- 0.00000E+00
7 2 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
8 2 0 raymond norm 0.837261E-02 +/- 0.20744E-03
9 3 1 wabs nH 10^22 0.000000E+00 +/- 0.11808E-01
Current statistic= 250.6 using 84 PHA bins.
XSPEC> p
!! Notice from the plot (which shows both the data and the latest model
spectrum) and the poor chi2 statistic that this is
!! a less than adequate fit, particularly in the higher channels.
XSPEC> recor
!! The recornrm command is used to renormalize the background correction
spectrum by a single multiplicative factor.
!! The final value of this correction should be between 0 and ~1.2; if the
correction lies outside this range of values, the
!! renormalization can be manually reset to a specified value using the cornorm
command. Alternatively, the user might
!! evaluate the quality of the data and/or the applicability of the adopted
model and, if necessary, restart with a different
!!model.
File # Correction
1 0.4263 +/- 0.0497
After correction norm adjustment 0.426 +/- 0.050 Chisquared = 177.05
!! Notice the resultant improvement in the chi2 statistic. Now
repeat the model fitting procedure.
XSPEC> fit 100
RENORM: renorm = 0.9176
Chi-squared Lvl Fit param # 1 2 3 4
5 6 7 8 9
Fit parameter 9 has pegged
146.67 0 5.820 1.000 0.0000E+00 1.3202E-02
0.7777 1.000 0.0000E+00 7.6068E-03 0.0000E+00
144.83 -1 4.791 1.000 0.0000E+00 1.2992E-02
0.7750 1.000 0.0000E+00 7.7216E-03 0.0000E+00
144.81 -2 4.710 1.000 0.0000E+00 1.2912E-02
0.7762 1.000 0.0000E+00 7.7649E-03 0.0000E+00
144.78 -3 4.625 1.000 0.0000E+00 1.2919E-02
0.7758 1.000 0.0000E+00 7.7489E-03 0.0000E+00
144.75 -4 4.460 1.000 0.0000E+00 1.3018E-02
0.7738 1.000 0.0000E+00 7.6718E-03 0.0000E+00
144.75 -5 4.461 1.000 0.0000E+00 1.3018E-02
0.7738 1.000 0.0000E+00 7.6717E-03 0.0000E+00
Unpegged 1 parameters
144.75 1 4.461 1.000 0.0000E+00 1.3018E-02
0.7738 1.000 0.0000E+00 7.6717E-03 0.0000E+00
1 1 1 raymond kT(keV) 4.46068 +/- 0.72606
2 1 2 raymond Abundanc 1.00000 +/- 0.00000E+00
3 1 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
4 1 0 raymond norm 0.130184E-01 +/- 0.43194E-03
5 2 1 raymond kT(keV) 0.773822 +/- 0.10575E-01
6 2 2 raymond Abundanc 1.00000 +/- 0.00000E+00
7 2 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
8 2 0 raymond norm 0.767169E-02 +/- 0.20749E-03
9 3 1 wabs nH 10^22 0.000000E+00 +/- 0.12717E-01
Current statistic= 144.7 using 84 PHA bins.
XSPEC> p
!! Again notice the improvement in the chi2 statistic. The fit is
now much improved compared to the previous iteration.
!! Now do another recor.
XSPEC> recor
File # Correction
1 0.6148 +/- 0.0497
After correction norm adjustment 1.442 +/- 0.117 Chisquared = 130.37
!! Notice that the chi2 value is better. Also, that the improvements
are getting smaller, i.e., the process appears to be
!!converging. Now do another couple of iterations of this 2-step process. But
first, notice that the interstellar column
!! density keeps pegging out at zero. Let us drop the wabs parameter from the model.
XSPEC> model raymond raymond
Input parameter value, delta, min, bot, top, and max values for ...
Fit parameter 1 of component 1 1 raymond kT(keV)
1.000 1.0000E-02 8.0800E-03 8.0800E-03 79.90 79.90
4.5
Fit parameter 2 of component 1 2 raymond Abundanc
1.000 -1.0000E-03 0.0000E+00 0.0000E+00 5.000 5.000
1
Fit parameter 3 of component 1 3 raymond Redshift
0.0000E+00 -1.0000E-03 0.0000E+00 0.0000E+00 2.000 2.000
0
Fit parameter 4 of component 1 0 raymond norm
1.000 1.0000E-03 0.0000E+00 0.0000E+00 1.0000E+05 1.0000E+06
0.013
Fit parameter 5 of component 2 1 raymond kT(keV)
1.000 1.0000E-02 8.0800E-03 8.0800E-03 79.90 79.90
0.77
Fit parameter 6 of component 2 2 raymond Abundanc
1.000 -1.0000E-03 0.0000E+00 0.0000E+00 5.000 5.000
1
Fit parameter 7 of component 2 3 raymond Redshift
0.0000E+00 -1.0000E-03 0.0000E+00 0.0000E+00 2.000 2.000
0
Fit parameter 8 of component 2 0 raymond norm
1.000 1.0000E-03 0.0000E+00 0.0000E+00 1.0000E+05 1.0000E+06
0.008
Summary of model parameters
Model Fit Model Component Parameter Value Flag
par par comp index type
1 1 1 1 raymond kT(keV) 4.50000 *
2 2 1 2 raymond Abundanc 1.00000
3 3 1 3 raymond Redshift 0.000000E+00
4 4 1 0 raymond norm 1.300000E-02*
5 5 2 1 raymond kT(keV) 0.770000 *
6 6 2 2 raymond Abundanc 1.00000
7 7 2 3 raymond Redshift 0.000000E+00
8 8 2 0 raymond norm 8.000000E-03*
4 variable fit parameters
Current statistic= 140.6 using 84 PHA bins.
XSPEC> fit 100
RENORM: renorm = 0.9362
Chi-squared Lvl Fit param # 1 2 3 4
5 6 7 8
124.11 -3 3.903 1.000 0.0000E+00 1.2378E-02
0.7743 1.000 0.0000E+00 7.3899E-03
124.11 -4 3.912 1.000 0.0000E+00 1.2380E-02
0.7743 1.000 0.0000E+00 7.3887E-03
1 1 1 raymond kT(keV) 3.91151 +/- 0.69366
2 1 2 raymond Abundanc 1.00000 +/- 0.00000E+00
3 1 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
4 1 0 raymond norm 0.123800E-01 +/- 0.43144E-03
5 2 1 raymond kT(keV) 0.774349 +/- 0.10978E-01
6 2 2 raymond Abundanc 1.00000 +/- 0.00000E+00
7 2 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
8 2 0 raymond norm 0.738867E-02 +/- 0.20767E-03
Current statistic= 124.1 using 84 PHA bins.
XSPEC> recor
File # Correction
1 0.6968 +/- 0.0497
After correction norm adjustment 1.133 +/- 0.081 Chisquared = 121.38
XSPEC> fit 100
RENORM: renorm = 0.9823
Chi-squared Lvl Fit param # 1 2 3 4
5 6 7 8
120.21 -3 3.656 1.000 0.0000E+00 1.2102E-02
0.7746 1.000 0.0000E+00 7.2654E-03
120.20 -4 3.703 1.000 0.0000E+00 1.2085E-02
0.7750 1.000 0.0000E+00 7.2808E-03
1 1 1 raymond kT(keV) 3.70303 +/- 0.45840
2 1 2 raymond Abundanc 1.00000 +/- 0.00000E+00
3 1 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
4 1 0 raymond norm 0.120845E-01 +/- 0.43102E-03
5 2 1 raymond kT(keV) 0.774994 +/- 0.11166E-01
6 2 2 raymond Abundanc 1.00000 +/- 0.00000E+00
7 2 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
8 2 0 raymond norm 0.728075E-02 +/- 0.20777E-03
Current statistic= 120.2 using 84 PHA bins.
XSPEC> recor
File # Correction
1 0.7322 +/- 0.0497
After correction norm adjustment 1.051 +/- 0.071 Chisquared = 119.69
XSPEC> fit 100
RENORM: renorm = 0.9922
Chi-squared Lvl Fit param # 1 2 3 4
5 6 7 8
119.47 -3 3.632 1.000 0.0000E+00 1.1963E-02
0.7750 1.000 0.0000E+00 7.2274E-03
119.47 -1 3.632 1.000 0.0000E+00 1.1963E-02
0.7750 1.000 0.0000E+00 7.2274E-03
1 1 1 raymond kT(keV) 3.63231 +/- 0.46753
2 1 2 raymond Abundanc 1.00000 +/- 0.00000E+00
3 1 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
4 1 0 raymond norm 0.119630E-01 +/- 0.43095E-03
5 2 1 raymond kT(keV) 0.775031 +/- 0.11225E-01
6 2 2 raymond Abundanc 1.00000 +/- 0.00000E+00
7 2 3 raymond Redshift 0.000000E+00 +/- 0.00000E+00
8 2 0 raymond norm 0.722745E-02 +/- 0.20793E-03
Current statistic= 119.5 using 84 PHA bins.
XSPEC> recor
File # Correction
1 0.7475 +/- 0.0497
After correction norm adjustment 1.021 +/- 0.068 Chisquared = 119.38
XSPEC> p
!! The fit is now fairly respectable. I have had enough, although purists may
want to converge even further.
XSPEC> cpd /ps
!! This sets my plot device so that the next time I plot a spectrum, it will
create a ps file that I can print as a PostScript
!! plot on my PostScript printer. This final plot of the data together with the
best-fit spectrum is shown below. Notice
!!that the final fit looks reasonable and has an acceptable chi2.
Figure 2
XSPEC> p
XSPEC> exit
Do you really want to exit (y) y
XSPEC: quit
!! This gives a flavor of how one might go about analyzing SSS spectra. The
particular models that are available in
!! XSPEC are discussed in the XSPEC User's Guide and in the on-line HELP facility.
The major points that the novice
!! should remember are that, unless you really know what you are doing, tell XSPEC to
ignore the bad channels in its
!! model fits, and that successful fitting is generally an iterative process involving a
number of repeats of fit and recor.
!! At present, if you want to do simultaneous SSS and MPC spectra modeling, you
have to enter XSPEC in its
!! stand-alone mode rather than interactively from BROWSE as in the above example. To do
this, one has to extract
!! the desired data products while in BROWSE, using the xp command.
SSS_TOTAL_CLA 34> xp/sp/assp 34
Extracting spectrum: 0.5-4.5 keV >> scapellaa.pha
Extracting associated spectrum: MPC spectra >> capellaa.pha
SSS_TOTAL_CLA 34> xp/li/asli 34
Extracting associated lightcurve: MPC spectra >> capellaa.rbf
Extracting lightcurve: 0.5-4.5 keV >> scapellaa.rbf
!! Extract SSS and MPC spectra and lightcurves for the Alf Aur (Capella)
observation on Day 61 of 1979.
SSS_TOTAL_CLA 34> exit
!! This ends the BROWSE session. In order to do detailed spectral analysis,
one must create background, correction
!! (instrument-dependent), and response matrix files .BCK, .COR, and .RSP
using VIMAT. Suppose we want to
!! model the SSS and MPC spectrum of Capella that we extracted using BROWSE as files
SCAPELLAA.PHA
!! and CAPELLAA.PHA, respectively. We must next type:
> vimat scapellaa
> vimat capellaa
!! and this will create the needed files for us. Now type XSPEC:
xspec 13:09:55 9-JAN-92
XSPEC> log
!! This makes a log file of the present XSPEC session. Notice that BROWSE makes
a log file automatically.
XSPEC> data scapellaa capellaa
!! This reads in all of the needed data files.
XSPEC> ignore bad
File 1 Ignored channels 1 to 1
File 1 Ignored channels 86 to 94
!! These channels are generally not recommended for use in model fits. Notice
that when XSPEC is called interactively
!! from BROWSE that this last command is done automatically. Now we can
proceed exactly as in the first example,
!! except that all of the fits will be to the combined SSS + MPC spectrum.
Conclusion
This gives the new user a brief glimpse of the power of the BROWSE and XSPEC
utilities to explore the SSS database resident at the HEASARC. But the only way
to really become familiar with our software and data is to get in there and
play with it! If you encounter any problems with the BROWSE or XSPEC programs,
or with the SSS data products, please contact us and we will provide any
necessary assistance.
Proceed to the next article
Return to the previous article
Select another article
HEASARC Home |
Observatories |
Archive |
Calibration |
Software |
Tools |
Students/Teachers/Public
Last modified: Monday, 19-Jun-2006 11:40:53 EDT
| 
