LEDAS  
Leicester Database and Archive Service
 
ARNIE Index
ARNIE Help
ARNIE Tutorial
Search...
All databases
All helpfiles
Line
Footer
E-mail

BATSE4B: Help

BATSE4B - CGRO/BATSE 4B Catalog


Overview

This database table comprises the 4th BATSE Gamma-Ray Burst Catalog, hereafter referred to as 4B. It specifies the locations and times for 1637 triggered gamma-ray bursts observed from 19 April, 1991 until 29 August, 1996. It therefore includes the data from the 3B catalog. The only revisions from the 3B catalog are improved locations for the trigger #s 741, 2311, and 3155.

Bursts since the end of the 1B catalog (March 1992) occurred when the Compton Gamma-Ray Observatory (CGRO) tape recorders were experiencing numerous errors. Consequently, there are gaps in the data of many bursts that preclude valid measurement of peak flux, peak rate, fluence, or duration. Peak rates on the 1 second timescale from each detector are almost always available. These data (called MAXBC rates) can be used to determine burst location. Previous difficulties with this data type have been largely removed, and we now believe that the systematic errors for MAXBC-located bursts are the same as for bursts located with other data types. It is still true however, that the MAXBC-located bursts usually have larger statistical errors than would be the case if another data type were available. The parameter called comments_position in this database contains comments on MAXBC-located bursts. A number of CGRO and BATSE flight software changes have significantly reduced the problem of data gaps since March of 1993.

Much additional material about the BATSE 4B Catalog can be found on the Compton Observatory Science Support Center website at

     http://cossc.gsfc.nasa.gov/docs/cgro/batse/4Bcatalog/
including some supporting tabular material: (i) An Exposure Table which describes the average exposure time as a function of declination for burst detection, (ii) a Trigger Efficiency Table which describes the fraction of time a burst will be triggered on each of the trigger timescales, as a function of the burst intensity, and (iii) a Trigger Criteria History Table which lists the trigger thresholds and the channels used for triggering as functions both of time and trigger number.

Catalog Bibcode

1998hgrb.symp....3M

References

The Fourth BATSE Gamma-Ray Burst Catalog, C. A. Meegan et al.,
    "Gamma-Ray Bursts: the Fourth Huntsville Symposium (IAP No. 428)",
    ed. Meegan, C.A., Preece, R.D. and Koshut, T.M., p. 1. (1997).

The Third BATSE BATSE Gamma-Ray Burst Catalog, C. A. Meegan et al.,
    ApJS, 106, 65. (1996).

Provenance

This database table was created by the HEASARC in March 1999, based on data supplied by the CGRO Science Support Center.

Description

The on-board software determines when a trigger occurs. When a burst trigger occurs, subsequent triggers are disabled for an accumulation period, during which the BATSE burst memories accumulate data. The accumulation period was 242 seconds until Dec 17, 1992, and from Jan 8, 1996 to Feb 25, 1997. At all other times it has been 573 seconds. The stored burst data are then transmitted; the readout time for all triggers was 90 minutes until Dec 17, 1992. At that date, the flight software was revised to suspend readouts during telemetry gaps and to truncate readouts of weak events. This resulted in a variable readout time. During the burst data readout, the 64 ms threshold is revised to correspond to the maximum rate attained by the current burst, and triggering is disabled on the 256 ms and 1024 ms timescales. Bursts intense enough to trigger over this revised 64 ms value are termed "overwrites". They appear as triggers in this catalog, with the overwrite flag is set to 'Y'.

The BATSE trigger numbers are the primary means of identification for events in this catalog. The trigger number is a running sequence of BATSE triggers which include cosmic bursts, solar flares and other events. The sequence begins with trigger 105 and ends with trigger 5586.

Each burst has a unique catalog name. These BATSE catalog names later may be incorporated into a multi-spacecraft catalog with "GRB" replacing this designation of "4B". The characters "4B " begin every BATSE catalog burst name, followed by the "yymmdd" of the burst. "yymmdd" is the two digit year, two digit month, and two digit day. When more than one gamma-ray burst occurs on one day, those bursts have a single letter suffix (B,C,D...), generally in order of intensity. Example: 4B 920504B refers to the second brightest burst that triggered BATSE May 3, 1992. The brightest burst on that day will have no suffix. The burst trigger time is the end of the interval (64, 256 or 1024 ms) on which the burst triggered the detector.

The angular error in position is the radius of a circle having the same area as the 68% confidence ellipse defined by the formal covariance matrix from a chi^2 fit on the assumption of normal errors. The error is based solely on the Poisson uncertainty in the BATSE measurement of burst flux by each Large Area Detector. There is, in addition, an RMS systematic error of approximately 1.6 degrees. Adding 1.6 degrees in quadrature to the error in the database table (the parameter called error_radius) yields the BATSE team's estimate of the 68% confidence interval for the burst location error. The statistical error is believed to be Gaussian.


Data Products

The following is a summary of the most useful BATSE data files and their contents:
Naming Convention         Contents

tte_bfits_YYYY.fits       time-sequenced 4-energy channel data bracketing
                          trigger; combines discla, preb, discsc, and tte.

discsc_bfits_YYYY.fits    time-sequenced 4-energy channel data bracketing
                          time of burst trigger for triggered detectors;
                          combines data types discla, preb and discsc.

(s)her_bfits_X_YYYY.fits  time-sequenced 128 energy channel data bracketing
                          time of burst trigger for specified detector;
                          combines datatypes her and herb - suggested data
                          type for 128 energy channel burst analysis.

mer_bfits_YYYY.fits       time-sequenced 16-energy channel data for times
                          bracketing burst trigger - triggered detectors
                          only; combines mer and cont datatypes - suggested
                          for 16-energy channel burst analysis.

discsc_drm_YYYY.fits      detector response matrix for 4 energy channel burst
                          data triggered detectors only; used with 4-energy
                          channel data to determine burst photon spectra.

her_drm_X_YYYY.fits       detector response matrix for 128-energy channel
                          burst data for specified detector; used with
                          128-energy channel counts data to generate burst
                          photon spectra.

mer_drm_YYYY.fits         detector response matrix for 16-energy channel burst
                          data triggered detectors only; used with 16-energy
                          channel data to generate burst photon spectra.

cont_TJD_fits             8 BATSE detectors, 2.048s resolution/16 energy
                          channels.

discla_TJD.fits           8 BATSE detectors, 1.024 s resolution/4 energy
                          channels.

XXX_TJD1_TJD2_his.fits    Occultation histories.

XXX_TJD1_TJD2_nhis.fits   Data for one or more BATSE detectors for available
                          energy channels as source count rate (counts/sec)
                          from which the background has been subtracted -
                          used for light curves.

XXX_TJD_lad_p11.fits      Pulsar low level data.

XXX_TJD1_TJD2_lad_olc     Light curve file for a given pulsar.
where YYYY = trigger number, XXXX = source name, TJD = Truncated Julian Day, X = detector number.

Notice that, because of the instrument configuration at the time of the event, the same files are not available for all triggers. Spectral (SD) data are prefixed with an 's' (e.g. 'sher').

Available data taken prior to the trigger may contain the beginning of the triggering event before it satisfied the triggering criteria. Background-type files can be used to remove background signal levels from the triggered period. The BFITS data files - containing burst and background spectral data as a function of time - and the detector response matrices (DRM) - modeling the instrument response to account for scattering and other effects - are extremely useful for gamma-ray burst analysis. Also, the BFITS and DRM files can be converted to PHA-II and RMF format for analysis with XSPEC using available FTOOLS. Other file types exist, notice, and advice on their use is obtainable at grohelp@athena.gsfc.nasa.gov.


Burst Durations

This database contains values for T90 and T50, quantities related to burst duration, for 1234 gamma-ray bursts that triggered the BATSE LAD detectors between April 1991 and 29 August 1996. T90 measures the duration of the time interval during which 90% of the total observed counts have been detected. The start of the T90 interval is defined by the time at which 5% of the total counts have been detected, and the end of the T90 interval is defined by the time at which 95% of the total counts have been detected. T50 is similarly defined using the times at which 25% and 75% of the counts have been detected. T90 and T50 are calculated using data summed over the 4 LAD discriminator channels and using data summed over only those detectors that satisfied the BATSE trigger criteria.

Users must note that T90 and T50 are not available for those bursts which suffer from data gaps during the event; the integration procedure inherently fails in these cases. However, visual estimates of the burst duration are provided in the parameter Comments_Duration for those bursts with sufficient data coverage. Users may also find other pertinent comments concerning the calculated value of T90 and T50 therein, and it is highly recommended that this parameter be consulted before any distribution selected on T90 or T50 is used.

Notice that the duration measurements for trigger 148 were recalculated after errors in the 3B values were brought to the BATSE team's attention by Dr. Jay Norris.


Count Rates

This database contains 912 triggered bursts observed from launch until 29 August 1996 for which peak count rates in units of the threshold count rate, and threshold count rates, are available. Many bursts do not have this information, particularly for those after March 1992, since insufficient data exist to determine either their peak counts or their threshold.

The BATSE on-board software tests for bursts by comparing the count rates on the eight large-area detectors to threshold levels for three separate time intervals: 64 ms, 256 ms, and 1024 ms. A burst trigger occurs if the count rate is above threshold in two or more detectors simultaneously. The thresholds are set by command to a specified number of standard deviations above back- ground (nominally 5.5 sigma). Background rates are recomputed every 17 seconds. The thresholds exhibit a coarse quantization that results from truncating the square root of the 64 ms count rate. Since we require that rates be above the thresholds of at least two detectors, the trigger threshold is determined by the threshold of the second most brightly illuminated detector. When a burst trigger occurs, subsequent triggers are disabled during the accumulation period when the BATSE burst memories accumulate data. These data are then transmitted. During this readout period, the 64 ms threshold is revised to correspond to the maximum rate attained by the current burst, and triggering is disabled on the 256 ms and 1024 ms timescales. Bursts intense enough to trigger during this readout period are termed "over- writes". They are recognized in the database by the value of -999 in the threshold values for 256 ms and 1024 ms (called threshold_256 and threshold_1024 in this database).

Since a trigger can occur on any of the three timescales, there are often cases in which the maximum rate will be below threshold on one or two of the timescales. The value of V/Vmax can be determined for any burst by selecting the maximum of the three peak rates, raised to the -3/2 power. Many bursts have unknown counts or thresholds on one or more timescales. These are marked by a "-999" in the database. This can happen for one of the follow- ing reasons: (1) If the trigger occurs on the 64 ms timescale during the peak 256 ms rate, then the peak 256 ms rate is not found; (2) If the 64 ms peak rate never exceeds the 64 ms threshold, and it occurs before the trigger time, then the peak 64 ms rate is not found; (3) If the 256 ms peak rate never exceeds the 256 ms threshold, and it occurs before the trigger time, then the peak 256 ms rate is not found. Note that items 2 and 3 do not affect V/Vmax, since these peak rates do not exceed threshold; item 1 can, on rare occasions, lead to an overestimate of V/Vmax.


Fluxes and Fluences

This database contains the fluence and peak flux values for the BATSE gamma-ray bursts between 19 April, 1991 and 29 August, 1996. There are 1292 bursts, each specified by the BATSE trigger number. All fluences and their errors have units of of ergs/cm^2. All peak fluxes and their errors have units of photons/cm^2/sec. The errors are the one-sigma statistical errors. The peak flux times are expressed in decimal seconds relative to the burst trigger time for the end of the interval in which the flux was calculated. The channel 1,2,3 and 4 fluences cover the energy ranges 20-50 keV, 50-100 keV, 100-300 keV, and E > 300 keV, respectively. The peak flux energy range is 50-300 keV, coinciding with the energy range of the nominal BATSE on-board burst trigger. Since channel 4 is an integral channel, fluences given for this channel are quite sensitive to the assumed spectral form. Spectral analyses in this energy range should be performed with higher resolution data types. Many of the bursts between March 1992 and March 1993 have significant gaps in the data and therefore do not have tabulated fluxes and fluences.

Parameters

Trigger_Num
The BATSE trigger number; this is a running sequence number of BATSE triggers which include cosmic bursts, solar flares and other events. The sequence in the 4B Catalog begins with trigger 105 and ends with trigger 5586.

Name
The BATSE Catalog burst name. Each burst has a unique catalog name. These BATSE catalog names later may be incorporated into a multi-spacecraft catalog with "GB" or "GRB" replacing this designation of "4B". The characters "4B " begin every 4B BATSE catalog burst name, followed by the "yymmdd" of the burst, where "yymmdd" is the two digit year, two digit month, and two digit day. When more than one gamma-ray burst occurs on one day, those bursts have a single letter suffix (B,C,D...) that is generally in order of intensity. Example: 4B 920503B refers to the second brightest burst that triggered BATSE on May 3, 1992. The brightest burst on that day will have no suffix.

RA
Right Ascension in the default equinox.

Dec
Declination in the default equinox.

LII
Galactic longitude in decimal degrees.

BII
Galactic latitude in decimal degrees.

Day_Trigger
The truncated Julian Date (TJD) of the trigger: TJD = JD - 2440000.5

Time
The burst trigger time, converted to shf for the BROWSE database or to MJD for the W3Browse database, for the convenience of searches and cross-correlations. The burst trigger time is the end of the interval (64, 256 or 1024 ms) in which the burst triggered the detector.

Seconds_Trigger
The burst trigger time, in decimal seconds of the day (UT) on which it occurred. The burst trigger time is the end of the interval (64, 256 or 1024 ms) in which the burst triggered the detector.

Error_Radius
The radius in decimal degrees of the positional error box. The error in angular location is the radius of a circle having the same area as the 68% confidence ellipse defined by the formal covariance matrix from a chi^2 fit on the assumption of normal errors. The error is based solely on the Poisson uncertainty in the BATSE measurement of burst flux by each Large Area Detector. There is, in addition, an RMS systematic error of approximately 1.6 degrees. Adding 1.6 degrees in quadrature to the value of this parameter yields the BATSE team's estimate of the 68% confidence interval for the burst location error. The statistical error is believed to be Gaussian. The systematic error distribution has a more extended tail than a Gaussian.

Earth_Angle
The angle in decimal degrees of the geocenter, i.e., the angle between the burst and the nadir, as measured from the satellite.

Overwrite
The overwrite flag: this is Y (true) if this burst overwrote an earlier, weaker trigger, N (false) otherwise.

Overwritten
The overwritten flag: this is Y (true) if this burst was overwritten by a later, more intense trigger, N (false) otherwise.

Max_Cts_64
The maximum counts in the second most brightly illuminated detector divided by the threshold count rate on the 64 ms timescale.

Threshold_64
The trigger threshold on the 64 ms time-scale. It is the number of counts in 64 ms required to trigger the second most brightly illuminated detector for this particular burst.

Flux_64
The peak flux on the 64 ms timescale in units of photons/cm^2/sec.

Flux_64_Error
The one-sigma statistical error in the peak flux on the 64 ms timescale.

Flux_64_Time
The time of the peak flux on the 64 ms timescale, in decimal seconds relative to the burst trigger time for the end of the interval in which the flux was calculated.

Max_Cts_256
The maximum counts in the second most brightly illuminated detector divided by the threshold count rate on the 256 ms timescale.

Threshold_256
The trigger threshold on the 256 ms timescale. It is the number of counts in 256 ms required to trigger the second most brightly illuminated detector for this particular burst.

Flux_256
The peak flux on the 256 ms timescale in units of photons/cm^2/sec.

Flux_256_Error
The one-sigma statistical error in the peak flux on the 256 ms timescale.

Flux_256_Time
The time of the peak flux on the 256 ms timescale, in decimal seconds relative to the burst trigger time for the end of the interval in which the flux was calculated.

Max_Cts_1024
The maximum counts in the second most brightly illuminated detector divided by the threshold count rate on the 1024 ms timescale.

Threshold_1024
The trigger threshold on the 1024 ms timescale. It is the number of counts in 1024 ms required to trigger the second most brightly illuminated detector for this particular burst.

Flux_1024
The peak flux on the 1024 ms timescale in units of photons/cm^2/sec.

Flux_1024_Error
The one-sigma statistical error in the peak flux on the 1024 ms timescale.

Flux_1024_Time
The time of the peak flux on the 1024 ms timescale, in decimal seconds relative to the burst trigger time for the end of the interval in which the flux was calculated.

T50
The 50% duration of the burst in seconds. T50 measures the duration of the time interval during which 50% of the total observed counts have been detected. The start of the T50 interval is defined by the time at which 25% of the total counts have been detected, and the end of the T50 interval is defined by the time at which 75% of the total counts have been detected.

T50_Error
The uncertainly in the T50 duration.

T50_Start
The start time of the T50 interval, relative to the trigger time (Time), in seconds.

T90
The 90% duration of the burst in seconds. T90 measures the duration of the time interval during which 90% of the total observed counts have been detected. The start of the T90 interval is defined by the time at which 5% of the total counts have been detected, and the end of the T90 interval is defined by the time at which 95% of the total counts have been detected.

T90_Error
The uncertainty in the T90 duration.

T90_Start
The start time of the T90 interval, relative to the trigger time (Time), in seconds.

Fluence_1
The fluence for Channel 1 (energy range 20-50 keV), in units of ergs/cm^2.

Fluence_1_Error
The error in the fluence for channel 1.

Fluence_2
The fluence for Channel 2 (energy range 50-100 keV), in units of ergs/cm^2.

Fluence_2_Error
The error in the fluence for channel 2.

Fluence_3
The fluence for Channel 3 (energy range 100-300 keV), in units of ergs/cm^2.

Fluence_3_Error
The error in the fluence for channel 3.

Fluence_4
The fluence for Channel 4 (energy range E > 300 keV), in units of ergs/cm^2. Since channel 4 is an integral channel, fluences given for this channel are quite sensitive to the assumed spectral form. Spectral analyses in this energy range should be performed with higher resolution data types.

Fluence_4_Error
The error in the fluence for channel 4.

Comments_Quality
Comments on data quality: not all gamma-ray bursts have such comments.

Comments_Otherobs
Comments on additional observations by other instruments: not all gamma-ray bursts have such comments.

Comments_General
General comments: not all gamma-ray bursts have such comments.

Comments_Position
Comments on the gamma-ray burst co-ordinates: not all gamma-ray bursts have such comments.

Comments_Duration
Comments on the gamma-ray burst duration: not all gamma-ray bursts have such comments.


Contact Person

Questions regarding the BATSE4B database table can be addressed to the HEASARC User Hotline.
If you have any problems, please consult the help page or mail ledas-help@star.le.ac.uk
 
HOME
SEARCH
SERVICES INFO SOFTWARE ViZieR BLASTA DSS
ARCHIVES ASCA CHANDRA GINGA ROSAT ARNIE