What is the status of the HRI spectral calibration?
At present, there is no plan for the US RSDC to develop a full scale response matrix for the HRI. Below is an excerpt from the HRI calibration report available on GSFC anonymous ftp node legacy.gsfc.nasa.gov, in the sub-directory /rosat/doc/hri/hri_report, discussing the spectral resolution of the HRI.
G. Fraser (1992, file spec_resp_zero.tex, on legacy.gsfc.nasa.gov, in the sub-directory /rosat/doc/hri) has analyzed the on-axis ground-based data by fitting Pearson functions to seven pulse height distributions. The relative FWHM of the fitted pulse height distributions vary from 66 percent (for the Al-K and Fe-K lines) to 107 percent (for the B-K line). Fraser concludes that "the average pulse height per pixel has very little dynamic range and is a very poor indicator of photon energy." However, he also notes that "the softness ratio", (counts in channels 1-5)/(counts in channels 6-11), used by (Wilson et.al. 1992, Ap.J. Lett., 391, L75) does seem the most sensitive energy indicator that can be constructed. (As noted below, the actual channels fulfilling this purpose will change as a function of the temporally-dependent gain.)
Although it is very difficult to obtain quantitative spectral information from the HRI (i.e., temperatures or power-law indices), it is still possible to obtain qualitative information on a scale of a few arcminutes using hardness (or softness) ratios. Wilson et. al. (1992, Ap.J. Lett., 391, L75) showed that the extended emission in NGC1068 is harder than the nuclear component using a softness ratio defined by the channels [1:5]/[6:11]. Since the HRI gain varies temporally and spatially, the user may want to define the PHA boundaries on a case-by-case basis. For example, in the two observations of N132D separated by two years, the fraction of the total counts in PHA channels 6-11 decreased from 40 percent to 10 percent.
Please use the Feedback link if you have questions on ROSAT.