AN UPDATED CALIBRATION OF THE ROSAT PSPC PARTICLE BACKGROUND FOR THE ANALYSIS OF DIFFUSE AND EXTENDED SOURCES

In order to permit quantitative studies of the cosmic diffuse X-ray ba ckground (DXRB) and of extended X-ray sources, we present updated cali brations of the particle-induced background of the Position Sensitive Proportional Counters (PSPCs) on board the Rontgen Satellite (ROSAT). We present new parameterizations of the temporal, spectral, and spatia l distributions of the particle-induced events following closely the a nalysis discussed in Snowden et al. (1992). The ROSAT Guest Observer ( GO) may find a step-by-step method for applying these parameterization s to a GO observation in sectional sign 3.4. Except for a variable con tamination which is present in channels less-than-or-equal-to 18 and a change in our understanding of the externally produced components, th e current parameterizations are quite similar to the previous results. We have used the spectral information available on the variable conta mination to formulate a method for determining the level of this conta mination in a given observation. The PSPC rejection efficiency for par ticle background events in the pulse-height range 18 less-than-or-equa l-to CH less-than-or-equal-to 249 is 99.90%, with a typical count rate of 4 x 10(-6) counts s-1 arcmin-2 keV-1. During typical conditions, t he count rate of residual events is well correlated with the Master Ve to count rate. The spectrum in the pulse-height range 18 less-than-or- equal-to CH less-than-or-equal-to 249 is well described by a power law , a flat component, and an Al Kalpha line at 1.5 keV. The spatial dist ribution of counts with pulse heights greater-than-or-equal-to 18 is u niform over the field of view except for a small radial gradient and s hadowing of the Al Kalpha line and part of the flat continuum by the w indow support structure. During an astronomical observation in low-gai n mode (after 1991 October 11), the particle background can also be mo nitored by the count rate in channels 260 less-than-or-equal-to CH les s-than-or-equal-to 370, since in most cases all these events are produ ced by particles. We have used a 54 ks observation of the Ursa Major r egion to verify the accuracy of our model. We have also presented a su mmary of the particle background parameterizations valid for four diff erent time epochs spanning the entire ROSAT mission, including a recal ibration of the data presented previously.