ROSAT Guest Observer Facility

ROSAT All Sky Survey Data: A Guide for Non-Expert users

The ROSAT All Sky Survey data were released to the public on March 11, 2000. This page describes the survey data which were processed at the Max-Planck-Institut fuer extraterrestrische Physik, how to access these data, and some points of concern.


The ROSAT All Sky Survey consists of 1378 distinct fields and was conducted in 4 distinct parts:
  1. A "mini-survey" was conducted during the performance verification checkout phase, from 1990 July 11 to 1990 July 16, using PSPC-C in scanning mode.

  2. The bulk of all sky survey was obtained by PSPC-C in scanning mode, starting on 1990 July 30 and lasting about 6 months. At GSFC, these data are archived using sequence identifications beginning with "RS9", for example

  3. With all but the last week of the six-month all-sky survey completed, an onboard computer glitch caused the ROSAT spacecraft to tumble out of control for approximately 15 hours. Housekeeping data taken during the tumbling showed that the satellite scanned across the sun, destroying PSPC-C. Portions of the final strip of the sky that went unobserved during this time was almost entirely surveyed by the PSPC-B in scanning mode between 1991 August 03-13. For these data, the accuracy of the attitude solution is somewhat worse since the WFC Startracker had to be used during this period.

  4. In February 1997 the PSPC-B was placed in the focal plan, and the final portion of the All Sky Survey was completed,this time in pointed mode, using PSPC-B. At GSFC, these pointed-mode "survey fill-in" data are archived using sequence identifications beginning with "RP19", for example:

Access to ROSAT All Sky Survey Data:

The ROSAT All Sky Survey Data was released to the general astronomical community by MPE in March 2000. These data are currently available from the MPE archive and the HEASARC archive at GSFC (at The All Sky Survey data is also incorporated into the public archive of ROSAT pointed data, and accessible by the standard means of access. Web access is also available from the MPE RASS page.

Some Notes about the Analysis of Scanning-Mode Survey Data:

While the pointed survey "fill-in" survey data (the "RP19" sequences) can be analyzed using the standard tools for analysis of pointed mod e data (such as PROS, EXSAS or the FTOOLS suite; for more info see the ROSAT Data Analysis page) the scanning mode data is a bit different than the more familiar pointed mode data, in these respects:

  • Each individual RASS field scanning-mode observation covers 6.4 x 6.4 degrees of sky, and are a combination of some number of individual PSPC fields (typically about 12 PSPC fields make up a single RASS field).

  • Important note about exposure times: The exposure time given in the FITS file headers by the ONTIME keyword represents the sum of all the accepted times for each field. In the case of pointing mode data, this number represents the exposure at the center of the field. However, since scanning mode data are comprised of a number of individual PSPC fields, the ONTIME keyword for scanning mode data represents the sum of the accepted times for all the individual PSPC fields comprising a given RASS field. Thus, for scanning mode data, the ONTIME value does not represent the exposure time of any given position in the RASS field. The actual exposure time for a position in a given RASS field can be obtained by looking at the MERGED EXPOSURE MAP (which has a name of the form rs9*_mex.fits). For example, for sequence rs931621n00, the ONTIME value is 1.502500E+04 seconds. By looking at the exposure map, we can see that no individual portion of this field received more than 443.9 seconds of exposure, and that the exposure time for this dataset varied between 117 and 443.9 seconds.

  • Currently neither PROS nor the FTOOLS suite is fully capable of handling the complete analysis of scanning mode data (though they can display and manipulate standard products such as images). Lightcurves and spectra can be extracted from the scanning mode data using FTOOLS, but these data products may not be directly usable (for example, spectra extracted from survey data using XSELECT will have incorrect exposure times assigned, since the exposure information relies on the value of the ONTIME keyword in the events file header). The ROSAT GOF in the US recommends that users interested in performing a full analysis of survey data MPE for guidance. MPE plans to include scanning mode analysis routines in the next public release of the MIDAS-EXSAS analysis package. See Procedures for Analysis of All-Sky Survey Data (Belloni, Hasinger and Izzo, 1994, A&A, 283, 1037) for a detailed discussion.

  • Some care needs to be taken when analyzing scanning mode survey spectra, since, in scanning mode, a given source will be observed at a wide variety of off-axis angles and the instrumental spectral response varies with off axis angle. This is not too disimilar to handling of pointed data, since even in pointed data a given source is observed at a variety of off-axis angles due to the spacecraft wobble. Spectral analysis of scanning mode data should use the appropriate on-axis PSPC-C response files (for unbinned 256-channel spectra, or for binned 34-channel spectra) and an ancillary response file which corrects the response for the off-axis histogram of the source photons.

Relevant Links

Page Author: Dr. Michael F. Corcoran
Please use the Feedback link if you have questions on ROSAT.

This file was last modified on Wednesday, 20-Oct-2021 11:04:40 EDT

NASA Astrophysics

  • FAQ/Comments/Feedback
  • Education Resources
  • Download Adobe Acrobat
  • A service of the Astrophysics Science Division (ASD) at NASA/ GSFC

    ROSAT Project Scientist: Dr. Robert Petre

    Responsible NASA Official: Phil Newman

    Privacy Policy and Important Notices.