HRC-S/LETG Grating Response Matrices [ vlk ; 2004dec03 ] ============================================================================== The CIAO tools MKGRMF and MKGARF allow creation of grating ARFs and RMFs for a specified dispersion order. However, because orders cannot be separated in the HRC-S, it is necessary to combine these ARFs and RMFs into a single response matrix in order that HRC-S/LETG spectra may be analyzed in XSPEC or Sherpa. The following files are the combined response matrices for orders 1 through 6, obtained by multiplying the CAL product gARFs (available from the LETG Observer Information page at http://asc.harvard.edu/cal/Links/Letg/User/Hrc_QE/ea_index.html ) with the RMFs currently in CALDB. rspdir/hrcslegp1.rsp -- gRSP(+1) rspdir/hrcslegp12.rsp -- gRSP(+1)+gRSP(+2) rspdir/hrcslegp1t3.rsp -- gRSP(+1)+gRSP(+2)+gRSP(+3) rspdir/hrcslegp1t4.rsp -- gRSP(+1)+gRSP(+2)+gRSP(+3)+gRSP(+4) rspdir/hrcslegp1t5.rsp -- gRSP(+1)+gRSP(+2)+gRSP(+3)+gRSP(+4)+gRSP(+5) rspdir/hrcslegp1t6.rsp -- gRSP(+1)+gRSP(+2)+gRSP(+3)+gRSP(+4)+gRSP(+5)+gRSP(+6) rspdir/hrcslegm1.rsp -- gRSP(-1) rspdir/hrcslegm12.rsp -- gRSP(-1)+gRSP(-2) rspdir/hrcslegm1t3.rsp -- gRSP(-1)+gRSP(-2)+gRSP(-3) rspdir/hrcslegm1t4.rsp -- gRSP(-1)+gRSP(-2)+gRSP(-3)+gRSP(-4) rspdir/hrcslegm1t5.rsp -- gRSP(-1)+gRSP(-2)+gRSP(-3)+gRSP(-4)+gRSP(-5) rspdir/hrcslegm1t6.rsp -- gRSP(-1)+gRSP(-2)+gRSP(-3)+gRSP(-4)+gRSP(-5)+gRSP(-6) where gRSP(order) == gARF(order)*gRMF(order) Notes:- 1. When using these gRSPs within, e.g., Sherpa, do NOT specify a gARF, since the gRSPs already include the gARFs. 1.1 For that matter, do not specify an EEARF either, because the gARFs that were used to create these gRSPs already include the correction factor due to the bow-tie extraction region. 1.2 Conversely, if you make gARFs using MKGARF and build your own gRSPs, you will have to use the EEARFs. 2. These responses have been tested extensively with continuum sources PKS 2155-304 and RX J1856.5-3754 and have been shown to produce reasonable results. 2.1 Best-fit values for PKS 2155-304 (ObsID 331) in the 10-100 Ang region (excluding the region of the plate gaps between 48-60 for +ve order and 58-70 for -ve order) are xwabs.nH = 0.0138977 10^22/cm^2 xspowerlaw.PhoIndx = 2.41015 xspowerlaw.norm = 0.0959052 photons/keV/cm**2/s at 1 keV which compares favorably with independent estimates of NH=(1.3-1.4)e20, PhoIndx=2.35-2.45, norm=0.09-0.1 (D.Pease, H.Marshall) 2.2 Best-fit values for RX J1856.5-3754 (full-res data from 1/2 Ms DDT observation) in the 12-100 Ang region, excluding the data gaps at 59-67 Ang for -ve order and 49-58 for +ve order, are xswabs.nH = 0.0110749 10^22/cm^2 xsbbody.kT = 0.0616318 keV xsbbody.norm = 0.000297661 L39/(D10)**2 which are close to independently determined estimates of NH=0.008-0.011, kT=0.059-0.064, norm=0.00027-0.00032 (A.Fruscione) 2.3 The gARFs are generated using corrections including orders 1-25. The gRSPs here only include 6 orders. Thus, the effective area at small wavelengths is underestimated (because of the missing contribution from the higher order gARFs), and power-law fits will therefore tend to overestimate the model intensity at small wavelengths to compensate. This generally causes a decrease of approximately 0.02 in the power-law index (e.g., PhoIndx for PKS 2155-304 ObsID 1704 is variously measured to be 2.42, but a fit with the response matrices given here results in a value of 2.40). 3. Care must be exercized in fitting line spectra using these response matrices, because of various limitations. e.g., 3.1 There exist shifts in the wavelength scale that are not necessarily accounted for in the RMFs, so fitting simultaneously to say a 1st order and 3rd order line may result in a poor fit for the latter because the predicted high-order location differs from the true location. 3.2 There also appear to be differences between the +ve and -ve order LSFs that are not (yet) included in the gRMFs, so simultaneously fitting +ve and -ve order lines is not recommended. 3.3 Beware of using the poly1d models for local continuum while fitting higher-order lines -- because there is no way to place wavelength limits on the continuum, the model predictions will include contributions from the lower-order also. 3.4 The same issue as above applies to independently modeling the background as power-laws (as recommended in the CIAO analysis threads). 3.5 The individual high-order gRMFs are stored at higher resolution than the 1st order gRMF. However, the final gRSP is perforce required to be on the same grid as the 1st order gRMF. Hence detailed information may be lost from the high-order response. If this detail is critical, you must use the individual gRMFs separately. 4. These response matrices were produced by sequentially adding gARFs and gRMFs of higher orders to that of the 1st order using IDL routines from PINTofALE ( http://hea-www.harvard.edu/PINTofALE/ ) and the IDLAstro library. Example code that may be used to derive hrcslegp1t3.rsp is given below, and may be generalized to any number of orders. 4.1 The example code uses canned gARFs, which are obviously not correct at plate gaps. If you use the output of MKGARF, note that that does not include the extraction efficiencies, so prior to forming the gRSPs, the gARFs must be multiplied by the appropriate extraction efficiencies. 4.2 The PINTofALE program ADDRSP() only adds two RMFs at a time; hence the need for the step-by-step approach detailed below. 4.3 The PINTofALE routine RD_OGIP_RMF() allows reading in the gRSPs and simultaneously reextract the effective area, as well as proxies for the effective area in channel space. 4.4 It is not necessary to install all of PINTofALE in order to use ADDRSP() and WRT_OGIP_RMF. It requires only the following programs, which are available online at http://hea-www.harvard.edu/PINTofALE/pro/util/ addrsp.pro kilroy.pro ogipzip.pro wrt_ogip_rmf.pro rd_ogip_rmf.pro rdarf.pro and also the IDLAstro library package, which is available at http://idlastro.gsfc.nasa.gov/homepage.html ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ; the following PINTofALE routines are required .compile rd_ogip_rmf rdarf addrsp wrt_ogip_rmf ogipzip kilroy ; where are the gRMFs? caldb=getenv('CALDB') rmfdir=caldb+'/data/chandra/tel/grating/letg/cpf/rmf/' ; gARFs may be downloaded from ; http://asc.harvard.edu/cal/Links/Letg/User/Hrc_QE/ea_index.html ; it is assumed that they are then placed in subdirectory arfdir/ ; read in 1st order gRMF and gARF rm1=rd_ogip_rmf(rmfdir+'/hrcleg1D1999-07-22rmfN0002.fits') ea1=rdarf('arfdir/letgs_EA_0411_o01p.fits',ostr1) ; interpolate the gARF to the grid of the gRMF ee=0.5*(rm1.elo+rm1.ehi) eaee=0.5*(ostr1.ELO+ostr1.EHI) & effar1=interpol(ea1,eaee,ee) > 0 ; read in 2nd order gRMF and gARF rm2=rd_ogip_rmf('rmfdir/hrcleg2D1999-07-22rmfN0002.fits') ea2=rdarf('arfdir/letgs_EA_0411_o02p.fits',ostr2) ; interpolate gARF to grid of gRMF ee=0.5*(rm2.elo+rm2.ehi) eaee=0.5*(ostr2.ELO+ostr2.EHI) & effar2=interpol(ea2,eaee,ee) > 0 ; combine 1st and 2nd orders rsp12=addrsp(rm1,rm2,ea1=effar1,ea2=effar2,verbose=10) ; read in 3rd order gRMF and gARF rm3=rd_ogip_rmf('rmfdir/hrcleg3D1999-07-22rmfN0002.fits') ea3=rdarf('arfdir/letgs_EA_0411_o03p.fits',ostr3) ; interpolate gARF to grid of gRMF ee=0.5*(rm3.elo+rm3.ehi) eaee=0.5*(ostr3.ELO+ostr3.EHI) & effar3=interpol(ea3,eaee,ee) > 0 ; add 3rd order to previously obtained combined 1st and 2nd order rsp1t3=addrsp(rsp12,rm3,ea2=effar3,verbose=10) ; write out to a disk file for use with Sherpa or XSPEC wrt_ogip_rmf,rsp12,'rspdir/hrcslegp12.rsp',telescop='CHANDRA',instrume='HRC',$ filter='NONE',detnam='HRC-S',grating='LETG',chantype='PI',$ cvsd='1999-07-22',cvst='00:00:00',mission='AXAF',$ hduclas3='FULL',creator='wrt_ogip_rmf.pro' wrt_ogip_rmf,rsp1t3,'rspdir/hrcslegp1t3.rsp',telescop='CHANDRA',instrume='HRC',$ filter='NONE',detnam='HRC-S',grating='LETG',chantype='PI',$ cvsd='1999-07-22',cvst='00:00:00',mission='AXAF',$ hduclas3='FULL',creator='wrt_ogip_rmf.pro' ============================================================================== [ history ] local file @ /data/snafu/kashyap/GRMF/RSP/README.GRSP Vinay L Kashyap 2001dec18 (vkashyap@cfa) updated gARFs (vlk 2002jan31) gARFs at 0.0125 AA bins (vlk 2002feb14) July 2002 gARFs (vlk 2002jul25) August 2004 gARFs (vlk 2004aug25) November 2004 gARFs (vlk 2004dec03)