ROSAT pointed observations have produced (and are still producing) an enormous amount of data. The PI of each pointing receives the output from the SASS including a source list and derived source results such as lightcurves and spectra; after one year the data (without the source lists or derived source data) are put into the ROSAT Data Archive (RDA) where they can be publicly accessed.
As a benefit to the astronomical user community, it was decided by the ROSAT project to put a significant effort into the production of reliable parameters for all the sources detected in the SASS processing of ROSAT pointed observations. The RRA is the result of the ROSAT project's effort to provide this source information.
The primary product associated with the RRA is a catalog of detected source parameters. This database may be accessed through the HEASARC BROWSE system, via any number of database engines, or as a simple ASCII table. The RRA catalog offers these improvements over existing catalogues of ROSAT sources (Voges et al. 1994, White et al. 1994):
The fact that ROSAT has very low background, excellent X-ray optics and a wide field of view means that ROSAT images of the X-ray sky are a complicated mix of large numbers of discrete sources within spatially-varying background emission. Additional complications arise from the detector structures (support structures or dead areas) and the spatially-variable point spread function of the ROSAT X-ray telescope and detectors. These complications mean that no automated source detection algorithm can be totally accurate, and that significant errors in source position, source significance and/or source intensity can occur in any such sample. Because of the large numbers of sources detected by ROSAT, manually extracting source parameters for all sources would be a difficult undertaking. The ROSAT project has adopted the following approach: we take the SASS output as a starting point and assign quality flags to the key parameters to alert the user to source parameters which are deemed suspicious or unreliable. Thus our philosophy is based on the premise that each user must ultimately take the responsibility to evaluate the reliability of a given parameter, but that the expertise of the ROSAT project may be helpful as a guide to the non-expert user of the ROSAT source data.
The flagging of detection parameters is done both automatically by a set of algorithms designed to use the combination of parameters to judge when a particular result is questionable, and manually by visual inspection. An example for the latter case is the determination that a detection falls between two close sources.
All source parameters that are suspect for any reason are flagged (e.g. detections which are close to the PSPC rib structure makes the position unreliable, or sources within extended emission where the source intensity is suspect). All SASS detections are checked so long as their S/N exceeds a minimum threshold below which it is difficult to distinguish real sources from statistical background fluctuations.
In summary, the RRA effort involves the construction of a pipeline (including human participation as one of its elements) which takes the SASS results as its input, and produces entries in a database, along with associated data products.