Chandra X-Ray Observatory
Skip to the navigation links
Last modified: 6 Jan 2016



CIAO 4.9 Science Threads



mkgarf creates a grating ARF for a particular order of an observation. The tool operates on a single detector element (i.e. one HRC plate) and grating order at a time. Therefore, it is necessary to generate all the individual component gARFs and then combine them with dmarfadd to create a grating ARF that covers the region of interest for that order.


To create grating ARFs for several orders of an LETG/HRC-S observation.

Related Links:

Last Update: 6 Jan 2016 - Review for CIAO 4.9. Only a single aspect histogram is needed since the GTIs for all the plates are the same.


About the Chandra Grating Data Archive and Catalog

The Chandra Grating Data Archive and Catalog (TGCat) is a browsable interface to analysis-quality spectral products (binned spectra and corresponding response files). TGCat makes it easy to find observations of a particular object, type of object, or type of observation, to quickly assess the quality and potential usefulness of the spectra with pre-computed graphics or custom-generated plots of binned and combined counts or fluxe spectra. Spectra, responses, event files, and summary products may be downloaded as a package.

TGCat runs standard CIAO tools, but also includes customized extractions for non-standard cases to refine the zeroth order position or to use regions appropriate for extended sources. Non-standard extractions details are provided in "verification and validation" comments for users.

Most public grating observations are available and new ones are added soon after they are released. See the list of of observations not included for exceptions. Many of the observations currently in this list will be included when we add enhanced processing for more difficult cases (multiple sources, very extended sources).

Please consider using the spectrum and responses (PHA, ARF, and RMF files) from TGCat in your analysis.

Background Information

mkgarf incorporates the telescope dither explicitly and averages over the response of the detector region which is exposed to a point in the sky. Hence, plate gaps and other detector non-uniformities, as well as the mirror vignetting function, are included. Arbitrary dither, off-axis response, or off-nominal source positions are not accounted for in the pre-computed effective area files.

The tool computes the efficiency for a full aperture in the cross-dispersion direction. The effect of the finite region width is calculated by mkgrmf, since the line-spread-function is physically coupled to the extraction width. This aperture efficiency factor is accounted for in analysis when using an ARF and RMF jointly to model the instrument. Sherpa, ISIS, and XSPEC will include this factor automatically when using both the ARF and RMF. ISIS also provides a factor_rsp function with which this energy-dependent term can be extracted, if it is explicitly needed.

Get Started

Download the sample data: 460 (LETG/HRC-S, 3C 273)

unix% download_chandra_obsid 460 evt2,pha2,asol,bpix,dtf
Early dataset

The data used in this thread was taken very early in the Chandra mission. It was not included in the bulk reprocessing metioned in the Watchout page. The keywords required for analysis with CIAO need to be added by running chandra_repro or r4_header_update.

It is assumed that you have created gRMFs for your observation by running the HRC Grating RMFs thread.

If you created a new bad pixel file by running the New Observation-Specific HRC Bad Pixel File thread, make sure that you have set up ardlib to use the same bad pixel file.

Can't I use fullgarf anymore?

A previous version of this thread simply used the fullgarf script to generate the gARFs. While the script correctly creates +/- 1 gARFS, the functionality does not yet exist to create higher orders.

For bright sources the contribution to the total flux from higher grating order dispersion will become significant. For those observations, it is recommended that users generate the gARFS "manually" by following this thread to include those additional orders.

Script It

The creation of the grating ARF and RMF is now automated.

chandra_repro will now automatically create the RMF and ARF for grating spectra. The results are stored in a tg/ subdirectory of the output directory.

Also, the mktgresp tool will create all the ARFs and RMFs for the grating spectra in a type II PHA file using the default grids and standard calibrations files. Just simply run

unix% mktgresp pha2 evt2 root_filename

where pha2 and evt2 are the filenames of your spectra output from tgextract or tgextract2 and the corresponding event file.

Thread Complete

If you have run chandra_repro or mktgresp then you are done with this thread and may move onto fitting and modeling your spectra. If you need to customize your responses or want to learn more details of the step by step processing steps you can continue with this thread.

Syntax note: foreach

This thread uses "foreach" loops to run the same CIAO tool for multiple chips. The syntax is correct for the csh or tcsh shell. If you are using another shell, e.g. bash, change the loop syntax accordingly.

Determine the orders to model

Because of the low energy resolution in the HRC-S, the PHA2 file contains two rows (negative and postive) (Figure 1) containing all the spectral orders. While it is not possible to separate the overlapping orders, CIAO can make response files (gRMFs and gARFs) for higher orders, allowing you to model and fit them.

We choose to focus on the first three positive and negative orders of the spectra. The Higher-order Responses for HRC-S/LETG Spectra thread provides some advice on determing the number of orders to model.

HRC-S plates and LETG orders

Three are three plates in the HRC detector: S1, S2, and S3. When the LETG grating is used, the orders are distributed as follows:

  • positive orders: S1 and S2
  • negative orders: S2 and S3

Further details on this are available from Table 9.3 in the The Proposer's Observatory Guide.

If you are only interested in the postive orders, for instance, it is unnecessary to create gARFs for the S3 plate; the files generated will contain no data. The "ARF was computed to be zero at all energies" Warning section contains some related information.

Compute the Aspect Histograms (asphist)

The aspect solution files are used to create a binned histogram detailing the aspect history of the observation. [Updated] HRC does not use separate good time intervals for each plate, therefore only a single aspect histogram is needed.

In some rare cases, there will be more than one aspect solution file (pcad_asol1.fits) for an observation. All the files must be input to the infile parameter, either as a list or a stack. Here we use:

unix% cat pcad_asol1.lis 

To create a histogram we run asphist once:

unix% punlearn asphist
unix% pset asphist infile=@pcad_asol1.lis
unix% pset asphist dtffile=hrcf00460_000N005_dtf1.fits
unix% pset asphist evtfile=hrcf00460N005_evt2.fits
unix% asphist outfile=460_asphist_s.fits mode=h

# asphist (CIAO): WARNING: skipping 41 livetime correction records
(from time: 63792195.736512 to time: 63792277.736515) 

The "skipping N livetime correction records" warnings are common and do not indicate a problem.

Get Source Position (dmlist)

The source position is required as one of the inputs to mkgarf. This information can be easily obtained from the PHA2 file with dmlist:

unix% dmlist hrcf00460N005_pha2.fits"[cols x,y]" opt=data
Data for Table Block SPECTRUM
ROW    X                    Y
     1          32830.43750     32641.6660156250
     2          32830.43750     32641.6660156250

The source is located at (32830.43750, 32641.6660156250).

Run mkgarf

Now we have all the information needed to run mkgarf. The tool needs to be run once per plate for each order; each order falls on two of the plates. Since we are interested in three positive and three negative orders, we need to run the tool twelve times:

  • S1: +1, +2, +3
  • S2: -1, -2, -3, +1, +2, +3
  • S3: -1, -2, -3

Again, refer to the HRC-S plates and LETG orders section if you need assistance determining how many times to run mkgarf. Remember that if you run it for the "wrong" plate, the tool will just issue a warning and create an empty file.

1. HRC-S1

First, we make a gARF for each order on plate S1. To simplify the steps, we loop through the three orders, using a variable to change the necessary parameters between runs. Setting the mode parameter to "h" tells the tool not to prompt for values.

Be sure that detsubsys=HRC-S1.

unix% punlearn mkgarf
unix% pset mkgarf asphistfile="460_asphist_s.fits[ASPHIST]"
unix% pset mkgarf obsfile="hrcf00460N005_evt2.fits[EVENTS]"
unix% pset mkgarf detsubsys=HRC-S1
unix% pset mkgarf grating_arm=LEG
unix% pset mkgarf sourcepixelx=32830.43750 sourcepixely=32641.6660156250

unix% foreach f ( 1 2 3 )
foreach? mkgarf outfile=garf_460_S1_LEG_${f}.fits order=$f \
         engrid="grid(460_leg_$f.grmf[cols ENERG_LO,ENERG_HI])" \
foreach? end

2. HRC-S2

The process is repeated for the S2 plate. Do not issue "punlearn" before running the tool again, as only a few parameters need to change. Note that the loop needs to cover the positive and negative orders, as they all fall on S2.

unix% pset mkgarf asphistfile="460_asphist_s.fits[ASPHIST]"
unix% pset mkgarf detsubsys=HRC-S2

unix% foreach f ( -1 -2 -3 1 2 3 )
foreach? mkgarf outfile=garf_460_S2_LEG_${f}.fits order=$f \
         engrid="grid(460_leg_$f.grmf[cols ENERG_LO,ENERG_HI])" \
foreach? end

3. HRC-S3

And, finally, gARFS for S3 are created.

unix% pset mkgarf asphistfile="460_asphist_s.fits[ASPHIST]"
unix% pset mkgarf detsubsys=HRC-S3

unix% foreach f ( -1 -2 -3 )
foreach? mkgarf outfile=garf_460_S3_LEG_${f}.fits order=$f \
         engrid="grid(460_leg_$f.grmf[cols ENERG_LO,ENERG_HI])" \
foreach? end

The history of each file, stored in the header, contains the complete mkgarf command used to create that gARF:

unix% dmhistory garf_460_S2_LEG_3.fits mkgarf
mkgarf asphistfile="460_asphist_s.fits[ASPHIST]"
outfile="garf_460_S2_LEG_3.fits" order="3" sourcepixelx="32830.4375"
sourcepixely="32641.666015625" engrid="grid(460_leg_3.grmf[cols
ENERG_LO,ENERG_HI])" obsfile="hrcf00460N005_evt2.fits[EVENTS]"
osipfile="CALDB" maskfile="NONE" mirror="hrma" detsubsys="HRC-S2"
grating_arm="LEG" pbkfile="" dafile="CALDB" ardlibparfile="ardlib.par"
geompar="geom" verbose="0" clobber="no"  

Combine the gARFs (dmarfadd)

dmarfadd is used to combine the individual gARFs for each order into a single response file. The tool is run six times, once for each order:

unix% punlearn dmarfadd
unix% dmarfadd infile="garf_460_S2_LEG_-1.fits,garf_460_S3_LEG_-1.fits" \

unix% dmarfadd infile="garf_460_S2_LEG_-2.fits,garf_460_S3_LEG_-2.fits" \

unix% dmarfadd infile="garf_460_S2_LEG_-3.fits,garf_460_S3_LEG_-3.fits" \

unix% dmarfadd infile="garf_460_S1_LEG_1.fits,garf_460_S2_LEG_1.fits" \

unix% dmarfadd infile="garf_460_S1_LEG_2.fits,garf_460_S2_LEG_2.fits" \

unix% dmarfadd infile="garf_460_S1_LEG_3.fits,garf_460_S2_LEG_3.fits" \


At this point, you should have the spectra, gARFs, and gRMFs necessary for fitting the data. The Fitting Multiple Orders of HRC-S/LETG Data thread shows how to load the data and responses, define a model, and fit the spectra.

In order to use Gaussian statistics to fit a model to a dataset, it is often necessary to "group" the data - i.e. combine channels until you have enough counts. Before fitting the data in Sherpa, read the Grouping a Grating Spectrum thread for more information.

The "ARF was computed to be zero at all energies" Warning

When running mkgarf, a warning of this form may be printed:

*** WARNING: The ARF was computed to be zero at all the specified energies.
             This is probably due to an incorrect source position

There are a few possible situations that are causing this:

  1. An incorrect source position was input to the sourcepixelx and sourcepixely parameters. Confirm that the source position is correct, and re-run the tool if it's not.

  2. You are attempting to create a gARF for an order that does not fall on that plate, e.g. a positive order on HRC-S3. Refer to the HRC-S plates and LETG orders section for more information.

If neither of these seems to fix the problem, contact the CXC Helpdesk for assistance.

Note on the Effective Area

The grating ARFs as generated by mkgarf are appropriate for events falling within the full 2pi steradian solid angle behind the grating. They do not include a correction for the fraction of these events that fall outside the "bowtie" extraction region whence the first-order spectra are selected from the L2 event file (tgextract). This fraction is of order 10%, and a correction for this is included in the RMF generated by mkgrmf, specifically in the LSFPARM calibration file.

Use of ARF and RMF files combined, e.g. within Sherpa, XSPEC, or ISIS, will result in the correct effective area being applied. Should they be required for analysis of LETG+HRC-S observations using methods or software that cannot easily combine RMF and ARF files, pre-generated on-axis LETG+HRC-S ARF files that include the bow-tie "extraction efficiency" are available on the LETG instrument team web page.


20 Dec 2004 updated for CIAO 3.2: canned gARFs are no longer available in the CALDB, users must run this thread, removed "Choosing an RMF" section
10 Feb 2005 new Calibration Updates entry for CALDB 3.0.1
08 Apr 2005 rewritten to no longer use the fullgarf script; added information on modeling higher orders
06 Dec 2005 updated for CIAO 3.3: new asphist tool syntax (the GTI filter is associated with the event file instead of the aspect solution file)
01 Dec 2006 updated for CIAO 3.4: CIAO version in warning
23 Jan 2008 updated for CIAO 4.0: parameter file updates for mkgarf (obsfile parameter is set to event file instead of aspect histogram file); Sherpa links point to 3.4 website; removed outdated calibration updates
12 Feb 2009 updated for CIAO 4.1: images are inline; sherpa link points to 4.1 website
19 Feb 2009 moved fitting information from Summary into Fitting section
16 Jun 2009 added About the Chandra Grating Data Archive and Catalog section
14 Jan 2009 reviewed for CIAO 4.2: no changes
05 Apr 2010 updated for CALDB 4.2.1: calibration update - the LETG grating efficiency has been upgraded to vN0006.
12 Jan 2011 reviewed for CIAO 4.3: no changes
26 Apr 2011 CALDB 4.4.3 release: calibration update - the LETG grating efficiency has been upgraded to vN0007.
30 Jun 2011 CALDB 4.4.5 release: calibration update - the HRC-S QEU has been upgraded to vN0006.
20 Jul 2011 required software updates are listed in Synopsis
10 Jan 2012 reviewed for CIAO 4.4: no changes
03 Dec 2012 Review for CIAO 4.5; no changes
24 Apr 2013 Add pointer to new mktgresp tool.
10 Dec 2013 Review for CIAO 4.6. Tried to clarify fullgarf/mktgresp usage.
22 Dec 2014 Review for CIAO 4.7. no changes
06 Jan 2016 Review for CIAO 4.9. Only a single aspect histogram is needed since the GTIs for all the plates are the same.

Last modified: 6 Jan 2016
Smithsonian Institute Smithsonian Institute

The Chandra X-Ray Center (CXC) is operated for NASA by the Smithsonian Astrophysical Observatory. 60 Garden Street, Cambridge, MA 02138 USA.   Email: Smithsonian Institution, Copyright © 1998-2016. All rights reserved.