Calibration status and caveats
- XIS mode: Normal & Full window, SCI-off, 5x5/3x3, Processing ver: 2.0, HEASOFT ver: 6.2, CALDB ver: **
- Time history of the center energy and energy resolution of the Mn Kalpha line
Figure 6 and 7 shows a time history (2005 August - 2007 April) of the apparent center energy and resolution of Mn Kalpha line from the 55Fe calibration source, which illuminate corners of the XIS0-3, segment A and D. For FI sensors, the gain are well calibrated untill July 2006 in FI, but afterwards gradually increased in XIS3 and rapidly decreased in XIS0 and XIS2. For BI sensor, gain are systematically higher in 10eV until July 2006, but gradually decreased afterwards. The energy resolutions are under 180eV untill July 2006 in both FI and BI sensors, but rapidly increased and became 300eV in April 2007.
Figure 6: Time history of the center energy of Mn Kalpha line for XIS0-3, segment A(left panels), D(right panels) for SCI-off Rev2.0 processed data. Each mark in the plots corresponds to an observation sequence.
Figure 7: Time history of the energy resolution of Mn Kalpha line for XIS0-3, segment A(left panels), D(right panels) for SCI-off Rev2.0 processed data.
- XIS mode: Normal & Full window, SCI-on, 5x5/3x3, Processing ver: 2.0, HEASOFT ver: 6.2, CALDB ver: **
- Time history of the center energy and energy resolution of the Mn Kalpha line
Figure 8 and 9 shows a time history (2006 August - 2007 April) of the apparent center energy and resolution of Mn Kalpha line from the 55Fe calibration source. During nine months, energy gain decreased about 15eV for FI sensors and 30eV dor BI sensor. The energy resolution are almost constant for FI sensors, but increased about 10eV for BI sensor during nine months.
Figure 8: Time history of the center energy of Mn Kalpha line for XIS0-3, segment A(left panels), D(right panels) for SCI-on Rev2.0 processed data. Keep in mind that the data are from August 2006 to April 2007, not same as figure 6.
Figure 9: Time history of the energy resolution of Mn Kalpha line for XIS0-3, segment A(left panels), D(right panels) for SCI-on Rev2.0 processed data.
- XIS mode: 1/4 window, SCI-off, 5x5/3x3, Processing ver: 2.x, HEASOFT ver: 6.2, CALDB ver: **
With the current calibration, the energy-scale of the 1/4 window mode has a negative offset to that of the normal mode. Those values are about -40 and -60 eV at 6 keV for the FI (XIS0,3) and the BI (XIS1) CCD, respectively.
- XIS mode: Normal, SCI-off, 2x2, Processing ver: 2.x, HEASOFT ver: 6.2, CALDB ver: **
- Energy scale
Figure 10 shows the energy-scale differences of the 2x2 editing mode from the 5x5 editing mode at ~1 keV (red) and ~6 keV (blue) as a function of the day since launch. Since the 2x2 editing mode is not used in BI, the XIS1 has not been investigated.
With the current calibration, the energy-scale of the 2x2 editing mode has a negative offset to that of the 5x5 editing mode. The offset is time-, energy-, and sensor-dependent while a significant positional dependency is not found.
Figure 10: Energy-scale differences of the 2x2 editing mode from the 5x5 editing mode.
- Energy resolution
Figure 11 shows the energy-resolution differences of the 2x2 editing mode from the 5x5 editing mode at ~1 keV (red) and ~6 keV (blue) as a function of the day since launch. Since the 2x2 editing mode is not used in BI, the XIS1 has not been investigated.
With the current calibration, a significant difference between the two editing modes is not found.
Figure 11: Energy-resolution differences of the 2x2 editing mode from the 5x5 editing mode.
- XIS mode: Normal, SCI-on, 5x5/3x3, Processing ver: 1.x, HEASOFT ver: 6.2, CALDB ver: **
- Time history of the center energy of the Mn Kalpha line (Ref. Provisional method to analyze the XIS data with SCI)
Figure 1 shows a time history (2006 August - 2007 February) of the apparent center energy of Mn Kalpha line from the 55Fe calibration source, which illuminate corners of the XIS0 and XIS1, segment A. Note that the PI channels are recalculated with a method described here. While the XIS1 shows a systematically lower energy scale of ~30 eV, the line center energies of the XIS0 are consistent with 5.895 keV (dashed-lines in the figures) within ~10 eV. See this page for more details.
Figure 1: Time history of the apparent center energy of Mn Kalpha line. The PI channels are recalculated with the zero-CTI CALDB files. Each mark in the plots corresponds to an observation sequence.
- Position dependence
Since charge loss due to the CTI depends on the distance from the readout nodes, it causes the ACTY-dependence of the pulse height, which should be properly corrected in order to perform precise measurements of X-ray energies. Although the CTI is very little for the SCI data, we investigated the ACTY dependence of the energy scale in the iron band and determined the CTI parameters so that the PHA channel of the Fe XXV Kalpha line from the Perseus cluster should become constant against ACTY. As shown in Figure 2, the PHA channels have no ACTY dependence when corrected for the CTI using the parameters given above.
For the SCI data processed with Ver 2, the PI channels are calculated by using the CTI parameters (see notes on the processing). However, the level of accuracy of the energy scale at various positions on the detector planes is under investigation. Thus users are recommended to check the energy scale of each sensor at the CCD corners by extracting the calibration sources' spectra.
Figure 2: CTI-corrected, PHA channel of Fe XXV Kalpha emission as a function of ACTY. The four panels correspond to the four segments of XIS3 and the green lines are the result of linear fitting to the data.
- XIS mode: Normal, SCI-off, 5x5/3x3, Processing ver: 1.x, HEASOFT ver: 6.2, CALDB ver:
- Time history of the center energy of the Mn Kalpha line
Figure 3 shows a time history (2006 June - 2006 December) of the apparent center energy of Mn Kalpha line from the 55Fe calibration source. Here the PI channels are calculated using the XISPI software and a calibration file "ae_xi[0-3]_makepi_2006_0823.fits"(*), and the line centroids are determined by fitting the 5.0-7.2keV PI spectra with the Gaussian functions. (*) The file incorporates the CTI parameters for the SCI-off data taken before 2007 July.
Figure 3: Time history of the center energy of Mn Kalpha line for XIS0-3, segment A, D. Earch mark in the plots corresponds to an observation sequence.
- Position dependence
Information on the position dependence of the energy scale for the SCI-off data is not currently available.
- Energy scale at low energies
We investigated the energy scales below 2 keV using the XIS, SCI-off data of a starburst galaxy M82 taken on 2005-October. The plots blow represent comparisons of center energies of three emission lines, He-like Mg at 1.34 keV, H-like Mg at 1.47 keV, and He-like Si line at 1.85 keV, derived with the Suzaku/XIS and the XMM-Newton/RGS. There is an good agreement between the two instruments within ~10 eV. However, users should be careful since inconsistency between the XIS and RGS energy scales on the order of ~30 eV has been found by the XIS team, for some of the SCI-off observations.
Figure 4: Center energies of low-energy emission lines derived with the four XIS sensors and the RGS onboard the XMM-Newton satellite. The upper left panel shows the RGS spectrum of M82 extracted from the archival data. The upper right, the bottom left, and the bottom right panels show the energies of He-like Mg, H-like Mg, He-like Si lines, respectively, where the red points are the results for XIS0-3 and the blue lines and the lightblue boxes indicate the best-fit line energies and their error ranges obtained from the RGS spectra.
- XIS mode: Window, SCI-on
Under investigation.
- XIS mode: Window, SCI-off
Under investigation.
- XIS mode: Burst
Under investigation.
- XIS mode: 2x2
- Energy scale at O8 and Ne9 lines
The energy scale for the 2x2 editing mode was investigated using the XIS0 data of E0102. As shown below, the gain difference between the 5x5 and 2x2 modes is kept constant ~7.5 eV.
Figure 5: Gain offset estimated at the 0.65 keV O8 Lyalpha line for the XIS0 data with the 2x2 editing mode. The energy offset [eV] relative to the 5x5 mode is plotted against the day since the launch.
Related links