Calibration Update

Larry David

The calibration team has used the ACIS external calibration source (ECS) to calibrate the ACIS gain ever since launch. ECS data is collected during normal operations before and after each radiation belt passage when ACIS is stowed. The ECS consists of three ⁵⁵Fe sources that produce Mn-Kα, Ti-Kα, and Al-Kα emission lines. ⁵⁵Fe has a half-life of 2.7 years, so the ECS has faded significantly over the 21 years of the Chandra mission. In addition, ACIS continues to operate at warmer temperatures as the mission progresses, while the fiducial point for ACIS gain calibration has been ECS data taken at cold temperatures (-120 < T < -119.2 C). Over the past year, the amount of cold ECS data acquired during normal operations (i.e., before and after radiation belt passages) has dropped significantly. This loss of cold ECS data required the scheduling of additional cold ECS data to be taken during science operations in 2020 (50 ksec in the first half of 2020 and 110 ksec in the second half of 2020) in order to obtain sufficient statistics to calibrate the ACIS gain. The time-dependent ACIS gain correction file (tgain) for the first half of 2020 was released in October 2020 and was based solely on cold ECS data. The tgain file for the second half of 2020 will also include warmer data (-120 < T < -117.2 C) to improve the statistics. The warmer ECS data requires some additional processing, but the resulting tgain file meets the calibration requirements on the ACIS gain (i.e., 68% of computed photon energies are within 0.3% of the lab energy). The tgain file for the last half of 2020 will be released in the next CALDB.

Due to the continued fading of the ECS, the calibration team is planning on switching to Cas A for future ACIS gain calibration. Cas A is a very bright thermal supernova remnant with many emission lines. The Cas A remnant is about 4^′ across, so multiple pointings would be required on each chip to fully calibrate the gain. However, a recent Principal Component Analysis (PCA) of the past tgain files shows that, for a given chip, there are persistent spatial structures in the tgain files that can be accurately reproduced by the first few components. Thus, it may be possible to just observe Cas A in the center of each chip and use the results of the PCA analysis to generate a tgain file for the entire chip. This is presently under investigation.

An updated version of the ACIS contamination file was released in December 2020. This version contains a slight adjustment to the build-up rate of the contaminant. Since this model was released, we have obtained data on Abell 1795, Mkn 421, and E0102. Analysis of these data show that the current model accurately predicts the continued build-up of molecular contamination on the ACIS filters.

The HRC experienced an anomaly in August 2020. Subsequent investigations concluded that the issue was with the side A electronics. In October to November, HRC operations resumed with a switch to the side B electronics. During the recovery of the HRC with the side B electronics, one of our primary calibration targets, AR Lac, was placed at the aim-point. The HRC-I and HRC-S observations of AR Lac with the side B electronics were nominal. Subsequent check-out observations of Cas A with the HRC-I and HZ43 with the HRC-S/LETG were also nominal. The observed count rate and gain in the HRC-S/LETG observation of HZ43 were consistent with the general rate of decline measured before the HRC anomaly. Following these check-out observations, normal science operations were resumed with the HRC. Additional calibration observations were carried out and an updated version of the HRC-S QE map was released on Dec. 16, 2020. Updated versions of the HRI-I QE file and the HRC-S gain file will be released in the next CALDB.

The HRC-I and HRC-S continue to experience a slow decline in gain and QE. This has necessitated a recent increase in the operating high voltage (HV) of both detectors. This is the first time the HV has been increased on the HRC-I and the second time for the HRC-S. A LETG/HRC-I observation of HZ43 was executed during the HV ramp-up of the HRC-I. Increasing the HV by two steps on the top and bottom plates restored the count rate to its value in 2017. The gain also increased significantly. A similar procedure was done with the HRC-S and the same two step increase in HV restored the HZ43 count rate to its 2015 level. Several calibration observations need to take place before new calibration products (e.g., gain and QE) can be developed with the new HV settings, but these should be available by the summer of 2021.

A paper on the “Concordance Model” has recently been submitted for publication. This paper presents a process for cross-calibrating the effective areas of X-ray telescopes that observe common targets. The targets are not assumed to be “standard candles” in the classic sense, but are assumed to have source fluxes that are well-defined, but a priori unknown values. Using a statistical method called shrinkage estimation, the effective area correction factors can be determined for each instrument that brings estimated fluxes into the best agreement, consistent with prior knowledge of their effective areas. The method is demonstrated with several data sets from various X-ray telescopes such as Chandra, XMM, Swift, and Suzaku.