TGCat processes all data within an automatic system to allow the best access to newly
public gratings data ready for analysis, and to allow the tgcat team to provide the best
extractions of gratings data with ease and little delay. All TGCat processing is done by
ISIS/S-Lang functions which set up and call CIAO tools.
grating/detector/mode combination supported:
Modes of ACIS:
TE: Timed Exposure
CC: Continuous Clocking
grating detector modes:
A quick guide to the extractions
Chandra data are identified by a "sequence number" and by an
"observation identifier", or OBSID. The sequence number generally
defines a single scientific observation, but due to operational
constraints, any given sequence number for a specific target may be
broken into several intervals, and each is assigned a unique OBSID.
We extract spectra by OBSID. During analysis, these may be
combined. In the summary plots, we show combined data in order to
provide the best overview of the quality of the observation. The
data products themselves maintain the individual components of the
spectra (different orders, gratings).
- In a later release, we intend to also provide data combined over
the obsids of each sequence number
Very briefly, the extraction process consists of the following
perform event processing (geometric and pulse-height
corrections; filtering on good events, good times);
determine the zeroth order's centroid (this is the
origin of the wavelength scale);
define a spatial mask which encloses spectral regions;
filter on the spatial mask and transform spatial to spectral
coordinates (wavelength, and cross-dispersion angle); if the
detctor is ACIS, use the CCD energy to resolve orders.
bin the spectra, sorted by order and grating, filtering on the
cross-dispersion region; from the region adjacent to the
source spectrum, bin a background spectrum (most important for
LETG/HRC, rarely important for HETG/ACIS).
Generate response files (effective area, or "ARF", and
line-spread-function, or "RMF");
compute a light curve from the diffracted photons;
compute some summary spectral quantities (counts and fluxes in
TGCat uses one of two automatic methods for determining the zero order position:
the standard CIAO tool, tgdetect
and for cases in which the zeroth order is blocked or heavily distorted by pileup, we use findzo, an ISIS program
which intersects the MEG trace with the ACIS readout streak.
The methods used are checked during the Validation and Verification ( V&V ). If the
method used failed to center on the zero order accurately, the alternate automatic method will be
used or the V&V scientist will provide either exact coordinates of the zero order centroid or a region
to act as an initial guess ( for cases where the target source is displaced from the proposers TARGET
All the extractions were inspected by a person familiar with the
instruments and spectra. We check that the zeroth order position is
accurate, that the spectral extraction region is appropriate, check
for avoidable source confusion, and that all the products exist.
The V&V procedure is guided by the same plots as presented in the
Comments and warnings are available in the results table ("v" link),
and as download products. Comments provide information about the
source which may be relevant for analysis. Warnings indicate
problems or features which may complicate analysis.
Some common comments and warnings are:
Comment: Bright, hard spectrum; extracted with narrow sky masks to
improve HEG flux at very short wavelengths (1.2-1.7A) (but at the
expense of background regions).
Comment: Mildly extended source <~size arcsec>
Comment: Complex source:
Comment: Multiple sources; extraction is of
Warning: Extended source, large extraction region used to include
Warning: Very extended source; special analysis required.
Warning: Background important
Warning: Spectra only valid in wavelength range [lo-hi] (energy
range [lo-hi]) due to detector masking.
Warning: Multiple sources in the field may cause contamination of
the primary extraction at some wavelengths.
Warning: CC-mode order-sorting with "osip=none" (a "flat" order
sorting PHA region) accepts some zeroth order scattering at short
Warning: ACIS T = -110C, no CTI correction available.
Warning: Very bright: spectra may be piled.
Warning: heavy "streaking" on S4 not rejected; (custom application
of "destreak" might mitigate).
VV process - Obsids Not in Catalog
There were some processing errors which caused us to reject
extraction until custom processing is done (
these are not included in the catalog):
Error: Multiple sources - background contaminated. Need to
extract only backgrounds from one side. [Mainly relevant for
LETG/HRC-S; default backgrounds are very wide.]
Error: Blocking of some orders/regions due to HRC-S shutters
(requires re-extraction of <+/-> side). (some Crab observations)
Error: far off-axis source - requires wider cross-dispersion
Finally, there were obsids which could not be processed due to inherent
limitations (e.g., HETG used with just the ACIS-S3 (acis-7) readout, or
LETG used with HRC-I); the list of these obsids can be viewed