|ROSAT Home Page||ROSAT
Figure 3.15: The ratio plot of data minus model residuals from a broken power-law fit to the PHA spectrum from Mkn 421. The absorbing column was allowed to be a free parameter (but the best-fit was close to the line-of-sight value). The plot illustrates the systematic deficit between 0.3-0.4keV observed in many spectra. The sharp rise above 2keV reflects the uncertainty in our knowledge of the optical constants used to calculate the effective area of the mirror.
For this post-launch matrix, a number of sources showed residual carbon absorption features which probably are an artifact of the calibration available at that time (see Fig. 3.15). After attempts at recalibration and after some alterations in detector operational parameters, it was decided that this matrix is currently the most appropriate to use for data taken before 1991 October 11 (S/C clock 42910000).
This corresponds to the following matrices: