1. Introduction and Summary
ascaarf v2.62, as well as the XRT response files, xrt_ea_v2_0.fits and xrt_psf_v2_0.fits, was released on July 1996 through ASCA GOF Previous versions of ascaarf (v2.53 and earlier) combined with older XRT responses (xrt_ea_v1_1.fits/xrt_psf_v1_1.fits and earlier), has been known to have the following problems:
These problems were corrected in ascaarf v2.62 and xrt_ea_v2_0.fits/xrt_psf_v2_0.fits. Essentially, two changes have been made:
Please note that ascaarf v2.62 has to be used with xrt_ea_v2_0.fits and xrt_psf_v2_0.fits to make the latest and most reliable ARFs. The correction filters are included in the ascaarf v.2.62 source code, and can be turned off by specifying the hidden parameter "arffil = no". By turning off the ARF fillter and combining with older XRT calibration files (xrt_ea_v1_1.fits and xrt_psf_v1_1.fits), ascaarf v2.62 is able to make the same old ARFs as were made with ascaarf v2.53 and ealier.
2. Direct Comparison of ARFs
Figure 1 shows ratios of ARFs made by ascaarf v2.62 (with xrt_ea_v2_0.fits and xrt_psf_v2_0.fits) and v2.53 (with xrt_ea_v1_1.fits and xrt_psf_v1_1.fits). The observation was 3C273 (June 9, 1993), and the point source method was used with the 4 and 6 arcmin integration radii for SIS and GIS respectively.
New GIS ARFs give 15 - 20 % smaller effective area values than old ones, hence GIS normalizations will become higher using the new ARFs. New SIS ARFs are also smaller than old ones by about 5 %. Overall, the ratio decreases from 2 keV toward 8 keV, resulting in slightly flatter spectral indices with the new ARFs. There are alo local structures with amplitudes of a few percents in the ratios.
Figure 1: Ratios of the ARFs made with ascaarf v2.62 (with xrt_ea_v2_0.fits and xrt_psf_v2_0.fits)
and v2.53 (with xrt_ea_v1_1.fits and xrt_psf_v1_1.fits) for the same
observation (3C273; June 9, 1993). Upper-left:SIS0, upper-right:SIS1,
lower-left:GIS2 and lower-right:GIS3.
3. Effect of the filters
As explained in section 1, ARF correction filters are introduced in the ascaarf v2.62 to absorb the normalization differences and local spectral structures the revised XRT response could not account for. GIS2 and 3 filters are made independently so that each Crab spectrum can be fitted with a power-law. Average of the two filters is used for both SIS, besides the normalization adjustment.
ascaarf v2.62 has an option not to apply this ARF filter to see its effect. In figure 2, we show ratios of the v2.62 ARFs (for 3C273 in June 1993) with the ARF filter to those without the filter.
Figure 2:
Ratios of the v2.62 ARFs with the ARF filters to those without the filters.
Upper-left:SIS0, upper-right:SIS1,
lower-left:GIS2 and lower-right:GIS3.
4. Consistency among the Four Detectors
3C273 (observed on June 9, 1993) and EXO2030+375 (observed on July 1, 1993) were used to check consistency of spectral parameters of the four sensors. The former source has a small low energy absorption, while the latter suffers from a significant absorption. In Table 1 - 4, we show results of model fitting for individual sensors (Table 1 and 3) and for four sensors simultaneously (Table 2 and 4), using ARFs with v2.53 and v2.62. The model for 3C273 is an absorbed power-law, and that for EXO2030 is an absorbed 'highecut'. Results of the model fitting for for 3C273 were shown in figure 3. In figure 5, we show results of simultaneous fitting for four sensors using ARFs with v2.53 and v2.62. GIS deadtime was corrected for EXO2030, which affects the normalization by 2 %.
With v2.53 ARFs, it is clear that GIS gives 20 % smaller normalizations than those with SIS. With v2.62 ARFs, GIS normalization increases by 20-30 %, and so does SIS normalization by 5 %. SIS and GIS normalizations are consistent within 5 % with the v2.62 ARFs.
5. Test with the Crab Nebula
The Crab Nebula was observed on September 28, 1994 for 35 ksec. It is known that GIS gain tends to become higher than real for bright sources such as Crab. Hence we adjusted the gain using the instrumental features at 2.2 keV and 4.8 keV as fiducial marks. Thus GIS2 and GIS3 gains were respectively made 99.4 % and 98.6 % of the original gain values determined in the REV1 processing. Deadtime correction is carried out (using the ftool 'deadtime').
In figure 6, results of the fitting with a power-law function are shown. It is seen that the residuals are less than 2.5 % with the v2.62 ARFs, although they are as much as 5 % with the v2.53 ARFs.
In Table 5, best-fit parameters are shown. Using the v2.62 ARFs, the 2 - 10 keV energy fluxes are 2.06 ergs s cm and 2.01 ergs s cm for GIS2 and GIS3 respectively. These are close enough to the standard value by Toors and Seward (Ast.J.79,995), 2.16 ergs s cm .
ARF version | Sensor | a | N b | Flux c | /dof (dof) | |
v2.53 | SIS0 | 2.9 1.5 | 2.72 0.07 | 1.65 0.02 | 3.46 | 1.66 (179) |
SIS1 | <0.5 | 2.60 | 1.65 | 3.40 | 1.25 (179) | |
GIS2 | <0.7 | 2.25 | 1.61 0.01 | 3.03 | 1.10 (94) | |
GIS3 | <1.2 | 2.37 | 1.63 | 3.14 | 1.34 (94) | |
v2.62 | SIS0 | <1.6 | 2.65 | 1.61 | 3.58 | 1.52 (179) |
SIS1 | <0.2 | 2.72 0.03 | 1.63 | 3.59 | 1.26 (179) | |
GIS2 | <0.5 | 2.69 | 1.58 | 3.71 | 1.11 (94) | |
GIS3 | <0.4 | 2.84 0.03 | 1.62 | 3.79 | 1.44 (94) |
a:10 cm .
b:10 photons s keV cm at 1 keV.
c:10 photons s cm for 0.9 - 10 keV.
Energy range is 0.9 - 10 keV.
ARF version | Sensor | a | N b | Flux c | /dof (dof) | |
v2.53 | < 0.07 | - | 1.625 | - | 1.41 (552) | |
SIS0 | - | 2.61 | - | 3.47 | - | |
SIS1 | - | 2.57 | - | 3.41 | - | |
GIS2 | - | 2.28 | - | 3.03 | - | |
GIS3 | - | 2.36 | - | 3.14 | - | |
v2.62 | < 0.01 | - | 1.610 | - | 1.40 (552) | |
SIS0 | - | 2.65 | - | 3.57 | - | |
SIS1 | - | 2.68 | - | 3.60 | - | |
GIS2 | - | 2.76 | - | 3.71 | - | |
GIS3 | - | 2.81 | - | 3.79 | - | |
a: 10 cm .
b: 10 photons s keV cm at 1 keV.
c: 10 photons s cm for 0.9 - 10 keV.
Energy range is 0.9 - 10 keV. Only normalizations are made free for individual sensors.
ARF version | Sensor | a | N b | Flux c | /dof (dof) | |||
v2.53 | SIS0 | 1.91 | 0.80 | 6.7 | 9.9 | 1.24 | 1.51 (179) | |
SIS1 | 1.91 | 0.83 | 6.6 | 18.5 | 1.24 | 1.66 (179) | ||
GIS2 | 1.81 | 0.82 | 6.4 | 11.0 | 1.05 | 1.17 (113) | ||
GIS3 | 1.91 | 0.91 | 6.5 | 9.6 | 1.05 | 1.24 (113) | ||
v2.62 | SIS0 | 1.91 | 0.77 | 6.8 | 9.0 | 1.30 | 1.38 (179) | |
SIS1 | 1.90 | 0.79 | 6.8 | 16.7 | 1.33 | 1.50 (179) | ||
GIS2 | 1.85 | 0.81 | 6.5 | 11.6 | 1.31 | 1.11 (113) | ||
GIS3 | 1.91 | 0.88 | 6.9 | 7.5 | 1.28 | 1.56 (113) |
a: 10 cm .
b: 10 photons s keV cm at 1 keV.
c: 10 photons s cm for 0.9 - 10 keV.
Energy range is 0.9 - 10 keV.
ARF version | Sensor | a | N b | Flux c | /dof (dof) | |||
v2.53 | - | 1.92 | - | 0.86 | 6.5 | 10.1 | - | 2.10 (592) |
SIS0 | - | - | - | - | 1.22 | - | ||
SIS1 | - | - | - | - | 1.21 | - | ||
GIS2 | - | - | - | - | 1.04 | - | ||
GIS3 | - | - | - | - | 1.06 | - | ||
v2.62 | - | 1.92 | - | 0.83 | 6.8 | 8.9 | - | 2.09 (592) |
SIS0 | - | - | - | - | 1.28 | - | ||
SIS1 | - | - | - | - | 1.30 | - | ||
GIS2 | - | - | - | - | 1.30 | - | ||
GIS3 | - | - | - | - | 1.29 | - |
a: 10 cm . b: 10 photons s keV cm at 1 keV. c: 10 photons s cm for 0.9 - 10 keV. Energy range is 0.9 - 10 keV. Only normalizations are made free for individual sensors.
a | Nb | Flux c | /dof (dof) | |||
v2.53 | ||||||
GIS2 | 3.1 | 8.00 | 2.13 | 1.65 | 4.41 (72) | |
GIS3 | 3.2 | 8.08 | 2.14 | 1.63 | 5.69 (72) | |
v2.62 | ||||||
GIS2 | 3.00 0.03 | 9.44 | 2.092 0.003 | 2.06 | 1.83 (72) | |
GIS3 | 2.70 0.04 | 9.07 0.04 | 2.083 0.003 | 2.01 | 1.64 (72) |
a: 10 cm .
b: photons s keV cm at 1 keV.
c: 10 ergs s cm for 2 - 10 keV.
Energy range is 0.9 - 10 keV.
Figure 3: Model fit for 3C273 with the v2.53 ARF (left) and v2.62 ARF (right) for
SIS0 (top) and SIS1 (bottom). SIS fit was carried out for 0.9-10 keV, but the spectra
were shown down to 0.5 keV for references. 3C273 may have a soft-excess component
over the power-law below 0.9 keV, and XRT/SIS calibration is less reliable
in this energy band.
Figure 4: Model fit for 3C273 with the v2.53 ARF (left) and v2.62 ARF (right) for
GIS2 (top) and GIS3 (bottom).
Figure 5: Unfolded spectra of four sensors simultaneously fitted.
Normalizations are allowed to
be free individually. Top two panels are for 3C273 and the bottom two panels are for EXO2030,
and v2.53 ARFs left and v2.62 ARFs right.
Two GIS spectra gave systematically lower normalizations than SIS with v2.53 ARFs,
but GIS and SIS normalizations are consistent with v2.62 ARFs.
Figure 6: Spectral fitting for GIS Crab spectra (top GIS2, bottom GIS3)
using the v2.53 ARF (left) and v2.62 ARF (right). Deadtime correction is carried out.
Next: About this document