Below are canned XRISM response files that can be used by GO Cycle 1 proposers to make their science technical feasibility. Detailed information on many important concepts related to these files are found in the Proposers' Observatory Guide (POG - available in html and pdf formats), which proposers are strongly encouraged to read before starting any feasibility study. Additional information on these response files, as well as examples of how to use them in Xspec, can be found in the presentation materials of the 2nd XRISM Community Workshop.
Redistribution Matrix Files (RMFs)
Three RMFs are available for Resolve while one RMF is available for Xtend. The Resolve RMFs are based on conservative estimates of high-resolution (Hp), mid-resolution (Mp), and low-resolution (Lp) primary events, whose relative fractions (a.k.a. the "branching ratios") depend primarily of the flux of the proposed source. For more information on branching ratios, we refer the proposers to the POG. Also note that the spectral resolution of all events degrade beyond ~6 keV. This effect remains marginal below ~8 keV, however it becomes important beyond 10 keV (about twice less good resolution at 12 keV).
Ancillary Response Files (ARFs)
Six ARFs are available for Resolve (extracted over the entire 35 pixels array) while two ARFs are available for Xtend (extracted over an on-aimpoint circular region of 5' radius). They are based on a few simple source cases (point-like or extended) seen through the telescope and the detector in various configurations (on- or off-aimpoint, with or without filter). The "aimpoint" here refers to the center of the detector (where most proposers will want their source to appear) and should not be confused with the axis position of its associated telescope. While these two points are virtually indistinguishable in Xtend, they are slightly offset in Resolve. Also note that all Resolve ARFs are provided assuming a closed Gate Valve (GV) only (see also Cycle 1 policies).
For each of the three ARFs provided for Resolve, the fraction of the input simulated photons ending up in the whole array (over the total number of photons in the focal plane) is:
Point source: 0.910
Flat circle: 0.232
Beta model: 0.467
For instance, proposers using a "pointsource" Resolve ARF to simulate point-like source at the center of the Resolve array should expect their source flux to be ~10% lower than initially modelled. This effect is already accounted for in the effective area calculation of the ARF.
Note that these reponse files are based on ground calibration only. In-flight calibration files will be available only from Cycle 2 (2025).
Note for extended source proposers
Using the "pointsource" Resolve ARF for extended sources results in a typical uncertainty on the modeled flux that can be up to to ~25%. In addition, the effective area curves in the Resolve "extbeta" and "extflatcircle" ARF files below are different from that of a point source because the incident photons enter the telescope with a range of off-axis angles, resulting in different efficiencies for impacting the focal-plane at a given energy. In spectral fitting, the ARF for an extended source gives a flux that corresponds to the full spatial extent of the source. For example, the "extflatcircle" ARF corresponds to a (flat) model flux extending out to 5' radius, hence beyond the Resolve array; its effective area is thus lower than the "pointsource" case because a smaller fraction of its input simulated photons have reached the array. Proposers of extended sources are thus encouraged to:
Either use the "extbeta" / "extflatcircle" ARF files and apply an appropriate rescaling of the flux/normalization of their source that accounts for this effect (e.g. depending on the size of their extracted flux region). By definition, this rescaling is 1 for a source emission with a 5' radius circle (5.7' radius circle) when using the "extflatcircle" ARF ("extbeta" ARF). This is recommended for sources whose surface brightness distribution resembles well one of the above two cases (i.e. nearly flat or distributed as beta model in a very similar fashion).
Or, perhaps more simply, use heasim (as well as its supporting files) and base their feasibility on the pointsource ARFs provided there. A quick tutorial is also available in the supporting files (see also quick video tutorial here).
This RMF assumes 100% of high-resolution events (Hp), conservatively estimated at 5 eV resolution in all pixels. For faint or
moderately bright sources, this RMF can be safely considered as
the standard RMF to be used for simulations. In case of bright sources,
however, more attention is required as the source counts might be
dominated by Mp - or even Lp - events and a more conservative RMF (see below) should
be used for feasibility.
This RMF assumes 100% of mid-resolution events (Mp), conservatively estimated at 6 eV resolution in all pixels. It should be used to simulate bright sources only (see POG for more details).
This RMF assumes 100% of low-resolution events (Lp), conservatively estimated at 18 eV resolution in all pixels. It should be used to simulate very bright sources only (see POG for more details).
This standard, baseline ARF can be used in most science cases. It assumes a
point source placed at the center of the detector array (i.e. on-
aimpoint but ~15'' offset from the telescope axis), with the filter
wheel set on the open position. The fraction of the input
simulated photons ending up in the detector area (over the total
number of photons in the focal plane) is 0.910.
This ARF assumes a point source, placed on-aimpoint, and
observed through the beryllium (Be) filter. This filter can be used to attenuate the flux of bright
sources and, thus, ensures a higher fraction of high-resolution events.
For more information on the use of filter on bright sources, proposers
are strongly encouraged to read the POG.
This ARF assumes a point source, placed on-aimpoint, and
observed through the neutral (ND) filter. This filter can be used to attenuate the flux of bright
sources and, thus, ensures a higher fraction of high-resolution events.
For more information on the use of filter on bright sources, proposers
are strongly encouraged to read the POG.
This ARF is based on the standard one (point source through with filter
wheel on open position), but with a ~1' offset compared to the on-
aimpoint position on the detector array. This can be useful for
proposers whose science case considers one (or several) source(s) to be
placed near the edge of the detector. Note that vignetting effects on
Resolve are expected to be limited, though non-negligible in some cases
(see POG).
This ARF assumes an on-aimpoint, open filter extended source distributed as a flat, circular
emission with a radius of 5'. The fraction of the input
simulated photons ending up in the detector area (over the total
number of photons in the focal plane) is 0.232. Since this ARF is extracted from all photons
originting from the flat circle model that made it onto the detector, its
normalization reflects directly this input modelled
region and should be adapted accordingly towards the flux of the proposed
extended source. For instance, the flux measured from a
spectrum simulated with this ARF corresponds to the flux from
that exact input region - i.e. a uniformly bright circular region of
5' radius.
We note, importantly, that this case is not representative of
reality, as observed extended sources (galaxy clusters, supernova
remnant, etc.) often show spatial and spectral complexity, resulting in multiple mixed spectral components. It is the proposers'
responsibility to adapt the feasibility of their proposed source
adequately.
Proposers are encouraged to consult the POG for more details on proposing extended sources.
This ARF assumes an on-aimpoint, open filter extended source distributed as a beta-
model with a core radius of 1.26', an outer extent of 5.7', and a
beta parameter of 0.53. The fraction of the input
simulated photons ending up in the detector area (over the total
number of photons in the focal plane) is 0.467. Since this ARF is extracted from all photons
originating from a circular beta-model that made it onto the detector, its
normalization reflects directly this input modelled
region and should be adapted accordingly towards the flux of the proposed
extended source (see, e.g., the "extflatcircle" example above).
We note, importantly, that this case is not representative of
reality, as observed extended sources (galaxy clusters, supernova
remnant, etc.) often show spatial and spectral complexity, resulting in multiple mixed spectral components. It is the proposers'
responsibility to adapt the feasibility of their proposed source
adequately.
Proposers are encouraged to consult the POG for more details on proposing extended sources.
Since Xtend is a CCD instrument, all photons are expected to be detected
with (approximately) similar spectral resolution regardless of the
brightness of the source. Hence, the same standard RMF can be used in
all simulations. Note, however, the risk of pileup for bright sources as
described further in the POG. Note: this RMF is different from the previously available file (xtd_standard.rmf, now deprecated). Although this version is extracted from older simulations, it better reproduces the spectral resolution of the actual Xtend instrument.
This standard, baseline ARF can be used in most science cases. It assumes a
point source placed at the center of the detector array (i.e. on-
aimpoint and on the telescope axis).
This ARF is also taken on aimpoint, but it assumes a flat circular
distribution (with radius of 5') instead of a single PSF. This ARF is
more appropriate to simulate extended sources. As for the Resolve
extended ARFs, we stress that this Xtend extended ARF does not reproduce
the spatial / spectral complexity of any known extended source. It is
the proposers' responsibility to adapt the feasibility of their proposed
source adequately.
Proposers are encouraged to consult the POG for more details on proposing extended sources.
This non-X-ray background spectrum was produced from night Earth observations with the Hitomi/SXI.
It is currently (March 2024) consistent with Xtend night Earth observations above 1 keV, but
the structure below 1 keV is less certain with cosmic ray echoes, which are under investigation.
It is extracted initially from a 2.5' radius circle at the aimpoint and multiplied by a factor
of 4 to match the above response files for a 5' radius circular extraction region.
Users may renormalize the spectrum to match their preferred
source extraction regions. This file does not include sky background
components, some of which may be significant for your target or science
and need to be evaluated to justify your proposed observation. Please
take a look at the instruction page on how to include the sky background
components to your simulation.
