|ROSAT Home Page||ROSAT
The processing of data in different bands (see Appendix B for a list of the defined pulse-height bands) is for the most part independent, and thus in the case where only a limited number of the broad pulse-height bands are of interest, it is in general unnecessary to process all of them. However, as noted below, the contamination by scattered solar X-rays in the R4 - R6 bands is linked since nearly all of the counts originate as O K emission at 0.53 keV (at the lower end of the R4 band). Counts are detected in the R5 and R6 bands only because of the limited energy resolution of the PSPC. A better lever on the solar contamination in the higher bands can be achieved by fitting the R4 band light curve, and the program which evaluates the contamination (TILT) requires the R4 band coefficients fitted by the program RATE_FIT.
Given the limited energy resolution of the PSPC, the most reasonable break point for excluding bands from analysis (if desired) is the distinction between the broad hard and soft bands. If the 1/4 keV band only is of interest, then the R1 (or R1L) and R2 bands only need to be processed. If the hard band is of interest, for instance a distant (and therefore absorbed) SNR in our galaxy, then only the R4-R7 bands need to be processed. The R3 band, since it has significant response on both sides of the carbon absorption edge of the PSPC window, is in general not useful and not supported.
In practice, it is quite often advantageous to combine bands for scientific analysis in order to improve count statistics for spatial analysis. The program FINAL_IMAGE allows this and combines the bands in a reasonable manner. However, all processing at the individual processing level (except for FINAL_IMAGE) must be run on all of the bands which are going to be combined into the larger band.