ASCA GIS Final Response

2005-05-09

ASCA GOF is going to release the final ASCA responses (XRT, SIS and GIS) as well as the final data processing (revision 3). In this page, I summarize the proposed ASCA GIS final responses.

Previous activites up to August 6 2001 are summarized at http://lheawww.gsfc.nasa.gov/users/ebisawa/GIS_NEW_RESPONSE/asca_new_response.html. See this page for the explanation of the files mentioned in the current page.

XRT responses

The following XRT responses should be adopted. These are created by running the ascaray ray-tracing code using the optical constants compiled by Keith Gendreau. Gold density is assumed to be 18.5 g/cm³.

kcg_v2_g18_5_xrt_ea.fits.gz

kcg_v2_g18_5_xrt_psf.fits.gz.

These files should be renamed, then put in CALDB and REFDATA. I would suggest names like xrt_ea_v3_0.fits and xrt_psf_v3_0.fits.

GIS RMFs

gis_off-5.0_pi500_l365.0_l220.0_l10.0_l1sl1.03_m525.0_m10.0_m23.0_m310.0_ebL30.995_eaL30.995.rmf

Of course, this should be renamed! I would suggest gis2v4_0.rmf and gis3v4_0.rmf for new names.

GIS ARFs

gisres.c

This function is found in the following location in the ftools package:

  $LHEASOFT/../src/ascalib/src/general/gisres.c

gisres.c file is used to create both GIS RMF and ARF (RMF builder is not released though). In the case of GIS, RMF efficiency is normalized to unity, and the detector efficiency is included in ARF. ascaarf requires gisres.c to calculate the GIS efficiency, as well as Be window transmission.

As of April 6, 2005, the release version is stamped Version 4.0, 1995-02-23.

GIS team modified this code and internally released gisres-v50test17gain1.0ofs.c in November 1999, in which Be and Xe absorption coefficeints are updated and Energy vs. pulse-hight relationship (including Xenon M and L edge jumps) are updated.

Later, parameters are adjusted at GSFC (energy scale, amount of edge jumps, Xe-thickness), and GIS team endorsed these changes. We propose to release this version of gisres.c (version 5), which will be the final one.

This version of gisres.c requires gis_rsp.h.

arffilter

Artificial arffilter is calculated in

  $LHEASOFT/../src/asca/src/ascaarf/arffil.c

However, the latest XRT and GIS response files do not require AC part of the arfilter. (DC part, normalization, may be required.) We need a mechanism to distinguish the old version of the data and responses (rev_2) and new version (rev_3), such that AC part of the arffilter is only applied to rev_2 datasets. If rev_2 and rev_3 response and data are mixed, ascaarf should abort the processing, or at least should give a strong warning.

Optical axis location and teldef files

Optical axis position is given in each teldef file with OPTAXISX and OPTAXISY keywords, which are given in mm.

Currently, the following OPTAXIS and OPTAXISY values are used:

gis2_ano_on_flf_180295.fits:

  OPTAXISX=           1.12500000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -0.61500000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

gis3_ano_on_flf_180295.fits:

  OPTAXISX=          -2.28500000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -1.48500000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

s0_teldef_ascalin.fits

  OPTAXISX=           0.60000000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=           2.20000000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

s1_teldef_ascalin.fits

  OPTAXISX=          -0.60000000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -3.60000000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

These values have to changed as follows (in mm) as new optical axis locations are determined on the detector coordinates. Suggested file names are also shown.

GIS2

  OPTAXISX=           0.93750000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -0.87500000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

GIS3

  OPTAXISX=          -2.06250000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -1.12500000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

SIS0

  OPTAXISX=           0.93150000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=           1.22850000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

SIS1

  OPTAXISX=          -0.21600000 / OPTICAL AXIS X OFFSET IN S/C COORD (MM)
  OPTAXISY=          -4.03650000 / OPTICAL AXIS Y OFFSET IN S/C COORD (MM)

ALL the teldef files in the CALDB have to be updated with these new keyword values.

ascalin

ascalin reads the teldef file OPTAXISX and OPTAXISY keyword values, convert millimeter to detector pixels, and write OPTIC1 and OPTIC2 keywords on the event file header. ascalin does not need any changes in introducing new RMF and ARF (but will need changes for taking account of other issues).

Just for ARF test, you may not have to run ascalin on the event file, but may modify the event file or spectral file header keywords:

SIS0:
OPTIC1 = 675.00
OPTIC2 = 595.00

SIS1:
OPTIC1 = 632.50
OPTIC2 = 790.00

GIS2:
OPTIC1 = 132.25
OPTIC2 = 132.00

GIS3:
OPTIC1 = 120.25
OPTIC2 = 133.00

ascaarf

ascaarf should use the new gisres.c and gisrsp.h mentioned above. arffil.c may need modification to introduce constant factors to adjust normalizations for the new responses. Also, as explaine above, some mechanism is needed to distinguish between the rev_2 data and calibration set and rev_3 set.

For the time being, I have built ascaarf without using the ASCA libraries distributed with ftools (i.e., I used the new gisres.c and gisrsp.h).

Creating ARF and consistency check with Crab

I have modified OPTIC1 and OPTIC2 values of g2.spec and g3.spec for Crab spectra.

ascaarf phafile=g2.spec rmffile=gis_off-5.0_pi500_l365.0_l220.0_l10.0_l1sl1.03_m525.0_m10.0_m23.0_m310.0_ebL30.995_eaL30.995.rmf outfile=g2.arf point=yes simple=yes fudge=no arffil=no xrtrsp=kcg_v2_g18_5_xrt_ea.fits xrtpsf=kcg_v2_g18_5_xrt_psf.fits

ascaarf phafile=g3.spec rmffile=gis_off-5.0_pi500_l365.0_l220.0_l10.0_l1sl1.03_m525.0_m10.0_m23.0_m310.0_ebL30.995_eaL30.995.rmf outfile=g3.arf point=yes simple=yes fudge=no arffil=no xrtrsp=kcg_v2_g18_5_xrt_ea.fits xrtpsf=kcg_v2_g18_5_xrt_psf.fits

GIS2 GIS3

g2.spec,g2.arf,xcm file, ps file g3.spec,g3.arf, xcm file, ps file

chi2 (dof) = 355 (192) chi2 (dof) = 424.6 (192)

Nh (1e22) = 0.361 Nh (1e22) = 0.378

alpha = 2.120 alpha = 2.143

N = 9.842 N = 9.793

flux (2-10 keV; erg/s/cm2) =2.047E-08 flux (2-10 keV; erg/s/cm2) = 1.964E-8

GIS2	GIS3
g2.spec,g2.arf,xcm file, ps file	g3.spec,g3.arf, xcm file, ps file
chi2 (dof) = 355 (192)	chi2 (dof) = 424.6 (192)
Nh (1e22) = 0.361	Nh (1e22) = 0.378
alpha = 2.120	alpha = 2.143
N = 9.842	N = 9.793
flux (2-10 keV; erg/s/cm2) =2.047E-08	flux (2-10 keV; erg/s/cm2) = 1.964E-8

Note that this is exactly the same as the result made on October 30, 2000, that must be the case.

Add 2 % systematic error (above 0.8 keV)

GIS2	GIS3
g2.spec,g2.arf,xcm file, ps file	g3.spec,g3.arf, xcm file, ps file
chi2 (dof) = 161 (192)	chi2 (dof) = 195 (192)
Nh (1e22) = 0.359	Nh (1e22) = 0.377
alpha = 2.120	alpha = 2.145
N = 9.836	N = 9.812
flux (2-10 keV; erg/s/cm2) =2.046E-08	flux (2-10 keV; erg/s/cm2) = 1.963E-8

Add 2 % systematic error (using down to 0.7 keV)

GIS2	GIS3
g2.spec,g2.arf,xcm file, ps file	g3.spec,g3.arf, xcm file, ps file
chi2 (dof) = 179 (195)	chi2 (dof) = 211 (195)
Nh (1e22) = 0.352	Nh (1e22) = 0.371
alpha = 2.116	alpha = 2.142
N = 9.777	N = 9.76
flux (2-10 keV; erg/s/cm2) =2.046E-08	flux (2-10 keV; erg/s/cm2) = 1.962E-8