ACIS-I3 Cycle 27 Aimpoint and Off-Axis RMFs for FP_TEMP>-120C
Due to the operation of ACIS at warmer focal-plane temperatures users extracting ACIS imaging spectra from line-rich, high-count sources could see line-centroids shift by 1-2% from systemic offsets due to calibration uncertainty. See Fig 1 below for an illustration of the effect on a simulated blackbody with a 1 keV gaussian line. Using RMFs for -120C (blue), -112C (orange), and -106C (green), the change in the ACIS-I3 response becomes clear.
Fig. 1: The effect of ACIS-I3 responses at three focal-plane temperatures
(see Table 1 below)
is illustrated
for a simulated blackbody spectrum with a gaussian line at 1.0 keV.
If proposers expect to have sources in the field of view of an ACIS observation that meet one of the following two conditions, they are encouraged to simulate spectra with the warmer response files (RMFs) in Table 1 below, and compare them to the standard -120 C RMF to determine if these focal plane temperatures will affect their proposed science.
- Spectra with more than 500 counts on a front-illuminated (FI) chip taken with ACIS focal-plane temperature between -105 C to -108 C.
- Spectra with more than 2500 counts on a front-illuminated (FI) chip taken with ACIS focal-plane temperature between -108 C to -111 C.
As with the previous set, these sample "warm" RMFs have been calculated using the standard fixed aimpoint positions for ACIS-I3 as the reference positions. The readout distances include that information. The corresponding ARFs, which don't depend on temperature, are included for convenience in the table below. The RMFs for -120C are numerically identical to the RMFs for the corresponding positions on the preceding ACIS RMFs/ARFs page.
A couple of useful "proposal planning" threads to implement the RMFs and ARFs in the table below:
- Simulating X-ray Spectral Data (PHA): the fake_pha command
-
Simulating Chandra ACIS-S Spectra with
Sherpa
Note this one specifies "ACIS-S", but ACIS-I RMFs and ARFs may be substituted.
To download individual files, <Shift>-click on the
filename. The complete set of responses is available as a
gzipped tarfile, located on the Chandra Data Archive server under the directory link:
https://cxc.cfa.harvard.edu/cdaftp/arcftp/ChandraCalDB/proposal/CY27/.
These files were built with CIAO 4.17 and CALDB 4.11.6. The RMFs are PI RMFs built with the CIAO phase-2 response tool mkacisrmf. They have 1024 linear energy bins (0.255 keV to 12.0 keV for both ACIS-I and ACIS-S), and 1024 linear PI bins (1-1024). (1-1024). All of the ACIS-I RMF files above are for the CTI-corrected response case, including the CTI corrections for ACIS-S3. That is, the P2_RESP input file used is acisD2000-01-29p2_respN0007.fits, and the CTI-corrected gain file acisD2000-01-29gain_ctiN0007.fits has been applied in mkacisrmf to set the PI energy scale.
The ARFs were built with mkarf, using the appropriate RMF for the energy scale. They are "2-pi" ARFs (no encircled energy correction) and assume a point source. They have been corrected for the time- and spatially-dependent contaminant absorption projected for the date 2026-05-15 (TT=895190469.184s), using the ARDLIB CONTAM file feature in CIAO 4.17, with the current ACIS contamination model, namely acisD1999-08-13contamN0015.fits. The N0015 CONTAM file was introduced in CalDB 4.11.6 on 2024-12-17, and contains a newly developed contamination model being introduced in Cycle 27.
Caveats
-
These ARFs and RMFs are for proposal planning purposes only. They should not be used for analysis of any real Chandra data. The Imaging Spectroscopy threads give instructions for building an observation-specific RMF and ARF.
-
The ACIS-I (FI) RMFs should be accurate to within a few percent between 0.5-6.0 keV. Below 0.5 keV, the effective area and gain for the FI devices is uncertain by 5-10%. Similar uncertainties apply above 6 keV. See the RESPONSE PRODUCTS RELEASE NOTES FOR ACIS for details.
-
The effective areas were derived for -120C ACIS focal plane temperature, and as such include CTI corrections in quantum efficiency valid for that temperature. These CTI corrections result in an energy-dependent QE gradient, which is visible in the effective area values at different distances from the readouts. Calculated count rates will be reasonably accurate without additional compensation for CTI when using these ARFs.