University of Birmingham

School of Physics and Astronomy

XMM-Newton EPIC Background Analysis

Please note: These pages and the software and files within them are no longer supported. A much more comprehensive study of the EPIC background is being undertaken by the EPIC Background Working Group (BGWG), and these pages, including newer software and files for the general user, are available from the official XMM-Newton site. The user is strongly recommended to visit this site.

XMM-Newton background Events files for the 3 EPIC instruments in their different instrument mode/filter combinations have been constructed using a superposition of many pointed observations. Background maps in several different instrument/mode/filter combinations and in several energy bands have also been constructed. On these pages, details can be found on how to obtain these background products together with related software and the paper on their construction and usage.


Latest updates : -

XMM-Newton background subtraction

When performing detailed XMM-Newton EPIC analysis, a good knowledge of the background is required. Sometimes it may be possible to extract the background from a region close to the particular source one is interested in (using a so-called `local' background). For a large or extended source however, one may have to extract the background far from the target source (the source may in fact be so extended, that no local background is visible within the field of view). Here, a number of effects can cause the extracted local background to be highly inappropriate in analysing the target source. If one believes that the use of a local background is impossible or will give rise to incorrect results, then the alternative is to use background data from a different observation, or better still, from several different observations. Hence the background products available here, which are described in more detail below.

Back to top

Available XMM-Newton background files

The Background Maps and Event files can be obtained from this ftp site.

Currently there are background events files available for the MOS1 and MOS2 in full-frame mode, and for the PN in both full-frame and full-frame-extended mode. Thin, medium and thick filter versions of these files are available for each of these instrument/mode combinations. Exposure maps are also available for each of these cases. For the thin and medium filters a set of background maps (photon maps and particle maps) are also available in a few predefined energy bands. In general the events files should be of more use to most people -- The events files are needed for spectral background subtraction, or for imaging analysis in a user selected energy band (this is also the only option for imaging analysis with the thick filter).

In the generation of the background events files and maps, sources detected in the standard SSC pipeline analysis of the constituent datasets were removed. Datasets with extended sources were also avoided.

The various files available are named as described in the explanatory notes, and the naming is also recapped below:

Back to top

Software available relating to background files

All related software can be obtained from the main background ftp site. As described in the explanatory notes, the user will usually work in sky coordinates (RA & Dec), not detector coordinates, and to a much finer scale than the 1 arcminute scale used in the production of the background maps. With this in mind, software has been developed to rebin and reproject onto the sky any of the above background maps to the spatial scale and sky position of a user-input image.

BGrebinimage2SKY is a shell script wrapper for the fortran routine BGrebinimage2SKY_interp. The shell script (BGrebinimage2SKY) and the fortran executable (either BGrebinimage2SKY_interp_solaris or BGrebinimage2SKY_interp_linux, depending on your system - this must be renamed to `BGrebinimage2SKY_interp') can again be obtained from the main background ftp site.

Both files, BGrebinimage2SKY & BGrebinimage2SKY_interp, should be placed in the user's directory (alternatively, one can edit the script so that it calls the fortran routine, wherever it is placed). One may need to change the protection of the fortran routine so that it is executable. Typing the name of the shell script wrapper BGrebinimage2SKY gives help on usage. See also the explanatory notes.

There is also (as of 29/05/03) a shell script wrapper to rebin and reproject the 4" resolution exposure maps. This is called BGrebinimage2SKY_4arcs, and can be obtained from the same ftp site. It uses the same BGrebinimage2SKY_interp (29/05/03 +) routine. The script is experimental, and because of the larger nature of the images involved, is very CPU-intensive, and may take some time.

There may also be some scripts of interest in the Background Analysis Scripts web page on this site (In particular the skycast script, also discussed briefly below).

Back to top

Production of Background Maps and Event Files

A paper describing this work has now been published:
The XMM-Newton EPIC Background: Production of Background Maps and Event Files    Read A.M., Ponman T.J., 2003, A&A, 409, 395

(The explanatory notes referred to in other parts of this document point to a local copy of this paper)

For the thick filter files (which were added after this paper was published) information on the original datasets making up these files can be found on the main ftp site in the file entitled 01_thick_filter_extra_notes.txt. The information in this file complements the information found in tables 3 and 4 of the main paper.

Back to top

Using these Background Files

A brief introduction

Reading the paper about these files first is probably a good idea, but a brief introduction to using these files is given below.

There are 2 sets of background products available: Background events files and background maps (which are images). In general the events files should be of more use to most people. The background images will be discussed first below.

Background maps (images)

The background images are 1 arcmin resolution maps in a number of different energy bands. These images are exposure corrected for any removed sources and in units of counts per second. The coordinate system used in these maps is detector coordinates (DETX/DETY). Usually, of course, the user will want to work in sky coordinates, not detector coordinates, and often the users data will be at a higher resolution than the 1 arcminute resolution of the maps.

To convert one of the background maps to a file appropriate for your observation use the script BGrebinimage2SKY (see the software section above). This program will rebin and reproject the 1 arcmin resolution detector images to the spatial scale and sky position of a user-input image. See the software section above and section 4.3 of the paper for a few more details.

The final product of this process will be an appropriate background image (photon map, particle map, or both combined) in the selected energy band, which is projected to the same resolution and sky position of your data.

Background Events files

Matching the events file to your data

To reproject one of the background events files to match your dataset, the script, skycast, can be used. This is available from, and described on the Background Analysis Scripts web page on this site. Basically this script will cast an XMM EPIC background dataset (or indeed any EPIC event dataset) onto the sky, at the position given by an input template event dataset or image.

Once the background events file has been reprojected to match the sky position of your data, one can extract either images or spectra from it. However, a quick look at any of the background events files will show partial holes in the data where sources were removed from the original files making up the combined background events file. To make any use of these backround files some allowance must be made for this effect.

Generating an exposure corrected image from an events file

The available exposure maps can be used to correct the uneven exposure time across an image extracted from the background events file.

NB - you may notice that for some of the exposure maps the peak values (units are seconds) are somewhat above the total livetime values given in table 4 of the paper (particularly for the PN). This is due to Out-of-Time Events (OoT) -- see the XMM-Newton Users' Handbook for more details on OoT events. Basically, these OoT events are events recorded while the ccd is being read out, so the total exposure time at any particular position is a combination of the total live time plus a fraction of time in which OoT events may be recorded.

The exposure maps provided are in DETX/DETY coordinates and are of 4 arcsec resolution. One possibility is that you could do all your work using this coordinate system and binning. Hence you could use the exposure maps directly to exposure-correct any (DETX/DETY 4") image created from the BG event files. This would be a bit restrictive though.

As listed above in the software section there is a shell script wrapper to rebin and reproject the 4" resolution exposure maps. This is called BGrebinimage2SKY_4arcs. It uses the same BGrebinimage2SKY_interp routine (as used by BGrebinimage2SKY). This script is still slightly experimental, and because of the larger nature of the images involved, is very CPU-intensive, and may take some time to run.

If any problems come up whilst using the BGrebinimage2SKY_4arcs script, another possibility is to rebin the exposure map images to a 1 arcmin resolution image and then use the much more robust BGrebinimage2SKY (described in the background images section above).

NB - BGrebinimage2SKY is currently hardwired to a 1 arcmin image resolution and BGrebinimage2SKY_4arcs is hardwired to a 4 arcsecond resolution image. It should be possible to write a generic, any resolution version, and the guts of it should be in these two scripts. It doesn't exist yet though.

Whichever route is taken, the end results should be an exposure map which matches the sky position and resolution of your dataset.

By this stage you should be able to take a background event dataset and reproject it to match the sky coordinates of your observation dataset (using skycast). You should also be able to generate an appropriate exposure map image for the background file (using BGrebinimage2SKY or BGrebinimage2SKY_4arcs). This should cover what is needed for an imaging analysis.

NB - You will want to use the events file to generate images rather than the background maps if you want to use an energy band which doesn't match any of the bands used for the generation of the background maps, or if you want to generate any sort of image for the thick filter (background maps don't exist for the thick filter).

Back to top

Further pages on XMM-Newton Data Analysis at Birmingham can be found here.

This page was produced by Andy Read and Stephen Helsdon
Andy Read - Home page

Last modified: