During a typical ROSAT observation, a number of spurious events are recognized as X-rays by the acquisition system. A classification scheme for this kind of event includes particle background (PB), scattered solar X-ray background (SB) and internal noise (IN). In the case of the ROSAT Position Sensitive Proportional Counter (PSPC, Pfeffermann & Briel 1986), a model exists for the contribution of PB and SB (Snowden et al. 1992, S92 hereafter; Snowden & Freyberg 1993, SF93 hereafter), while the internal noise is negligible. In the case of the HRI, a model exists for the IN (David et al. 1995) and PB (Snowden, in preparation), but the effect of SB has not been investigated in detail as for the PSPC case. In all cases, the standard procedure to reduce the effects of the contamination by non cosmic backgrounds involves, at a certain point, a "cut" of putative offending observation intervals. The choice of what and where to cut is crucial, since it should actually be an optimal trade-off between two effects: the decrease of science target counting statistics and the reduction of the contamination by spurious events.
Here we describe a procedure to screen out the high background intervals from X-ray observations done with the ROSAT PSPC or HRI. Our approach differs from the one in S92 and SF93, because of the following reasons. First, it is model independent, i.e. it does not try to model the origin of the background, but it operates on the observed shape of the background light curve, thus eliminating the dependency of the method on the input model parameters as well as on the validity of the model itself. Second, it is optimized in order to guarantee that the resulting screening maximizes the signal-to-noise ratio (SNR) of faint point-like sources in the field of view (FOV). Since detection algorithms operate typically at a fixed source acceptance probability threshold, i.e. at a given SNR, the application of the optimized screening procedure may yield a significant increase of the number of faint sources detected in the FOV.
Our paper is organized as follows: in Sect. 2 we describe the screening algorithm from a general point of view. In Sects. 3 and 4, we apply the formalism developed in Sect. 2 to the ROSAT PSPC and HRI cases and we draw some preliminary conclusions. In Sect. 5, we outline the operative recipe of the PSPC screening code we have developed under the IRAF environment, and in Sect. 6 we show the results of the code in the case of a PSPC image. Finally, in Sect. 7, we summarize our results.