X-Ray Background Tool Help


This tool calculates the average X-ray background count rate and statistical uncertainty in each of the six standard bands of the ROSAT All-Sky Survey diffuse background maps (R1, R2, R4, R5, R6, R7) for a specified astronomical position and either a circle with a specified radius or an annulus with specified inner and outer radii centered on this position . In addition, the average count rates are given for the combined 1/4 keV (R1 + R2), 3/4 keV (R4 + R5), and 1.5 keV (R6 + R7) bands. (The count rates for the combined bands are simply the sum of the count rates for each component band, and the uncertainty in the combined band count rate is the square root of the sum of the squares of the uncertainties of each component band.) The values returned by the tool are in units of 10-6 counts/second/arcminute2. The software calculates the average count rate using the map values within N degrees from the requested position.

The software starts by creating a sub-sample of each map with size M degrees X M degrees, where M is an internal parameter chosen to be a value slightly larger than either 5 degrees (if the specified cone radius is less than 1.25 degrees) or four times the specified cone radius (or outer radius of annulus) N, if N is 1.25 degrees or larger), up to a maximum of 20 degrees for a maximum possible radius of 10 degrees. Only the pixel map values within N degrees are used to calculate the average count rate value.

Note: The RASS X-ray data are taken from six map tiles centered on the Galactic plane at l=0, 90, 180, and 270 degrees, and around the north and south Galactic poles. They are ZEA (Zenith Equal Area) map projections, and the pixels are fairly square over the entire ZEA map. For any data extraction, data from only one map tile (file) is used. Since these map tiles have a minimum of 16 degrees of overlap, the largest regions (radii of 8-10 degrees) near the tile borders will not be fully sampled. In these cases the area outside of the source map tile is ignored.

Search Region

There are two possible search areas for calculating the X-ray background count rates for a specified astronomical position.

The Cone search option uses all pixels within the radius N of a circle centered on the astronomical position. This search region is the default.

The Annulus search option uses the pixels within a ring centered at the specified astronomical position, with outer radius R1 and inner radius R2, such that R2 is less than R1. As R2 approaches zero, the Annulus and Cone search areas become the same thing. This feature is useful for removing extended sources (e.g., a cluster of galaxies) from disturbing the calculation of the X-ray background count rates.

Radius and Annulus Radii

Only map values within the specified radius are used in the average count rate calculation. The default is 1 degree. The maximum allowed radius is 10 degrees. All radius values should be specified in units of degrees.

If the Annulus option is selected, the Radius field specifies the outer radius of the ring that defines the annulus region. The Inner Radius of the ring must also be specified, and it must be less than the outer radius. Inner radius values should also be specified in units of degrees.

Note: If the search radius is larger than the amount of overlap between map tiles (see Description above) and the chosen position is near the edge of the map being used, the computed averages may not include some values that would extend beyond the edge of the map tile.

Note: Increasing the radius will increase the likelihood that the average will be over true variations in the diffuse X-ray background. Decreasing the search radius may not provide sufficient coverage to allow for reasonable statistical uncertainties. (The survey was photon limited.) If the search radius is decreased to less than 0.2 degrees, the program may not find any data points within the specified radius. Use results with caution.

Plotting the X-Ray Background Average Count Rate with XIMAGE

Checking this option will plot an image of the X-ray background average count rate for a region around the requested position. Each pixel in the image is constructed using the counts, background counts, and exposures of the corresponding pixels from all seven bands of the ROSAT All-Sky Survey diffuse background maps. Each pixel is really the average computed using the corresponding map pixels and the eight pixels surrounding it, using the following formula for each image pixel, Ii j:
Ii j = 0.39063 * (sum(Ci j k) - sum(Bi j k)) / <E>

where the sums are over the pixel coordinates range [ i=i-1 to i+1, j=j-1 to j+1 ] and the band range [ k=1 to 7 ]. Ci j k represents the counts for band k at pixel (i,j) in the ROSAT All-Sky Survey diffuse background maps. Bi j k are the background counts  and Ei j k are the exposure times. Values of Ii j have units counts/s/arcmin2. The scale factor of 0.39063 is the magic number required to produce those units and comes from the fact that the pixels for the basic survey data arrays are 1.6 x 1.6 arcmin2. <E> is the exposure weight and is defined as

<E> = sum(Ck * Ek) / sum(Ck)

where the sums are over the band range [ k=1 to 7 ]. Ck and Ek are defined as follows:

Ck = sum(Ci j k)
Ek = sum(Ei j k)

where these sums are over only the pixel coordinates range [ i=i-1 to i+1, j=j-1 to j+1 ].

In the interest of providing visual context for the search region, the width and height of the image is four times the size of the cone (or outer annulus) radius, up to a maximum of 20 degrees, or 5 degrees, whichever is greater. Once the values of the pixels have been obtained and written to a FITS file (which also contains the spectrum), XIMAGE is used to plot the image for the Web interface. Cross-hairs are added to the image, showing the exact position specified. Green circles are also added, signifying the search region, two concentric circles for an annulus or one circle otherwise.

XIMAGE Plotting Option: Scaling

The Web interface has an option which allows the user to specify the scaling used by XIMAGE to color the image. The scaling used in displaying an image can dramatically affect its appearance. By default Histogram equalization is used, however Linear, Logarithmic, and Square-Root scaling are also available. The following mosaic shows the same image displayed using each of the scaling options:

Images of Abell 2142 comparing histogram equalization, linear, logarithmic, and square-root scaling

The best scaling to use depends on the character of the image data. Histogram equalization works well with most distributions, as it maps pixels into color bins such that each color is used to draw about the same number of pixels. For low count images, however, there are not enough unique pixel values to fill up the available colors. Linear scaling is useful for data values which are evenly distributed and with a relatively small range, while logarithmic and square-root scaling are better for data values with a large range.

XIMAGE Plotting Option: Overlay Contours

A contour plot will be overlaid onto the image if the Overlay Contours option is checked. By default the contours are calculated using the same scaling method specified for the image with the range divided into ten levels with the contours for level 5 and higher are plotted on top of the image.

Approximate Photon Energy Ranges for ROSAT Background Map Bands

Band PI Channels Approximate Energy Range (keV)
R1 8-19 0.11-0.284
R2 20-41 0.14-0.284
R12 8-41 0.12-0.284
R3 42-51 0.20-0.83
R4 52-69 0.44-1.01
R5 70-90 0.56-1.21
R45 52-90 0.47-1.21
R6 91-131 0.73-1.56
R7 132-201 1.05-2.04
R67 91-201 0.76-2.04

The energy of each band corresponds to 10 percent of the peak response. Bands with the same upper (or lower) channel boundaries can have different upper (or lower) energy range limits because of the different widths of the bands and the definition of the energy range as a fixed fraction of the peak response. (See Snowden et al., ApJ, Vol. 424 (1994), pp. 714-728, in the References section listed below).

Determining the Integration Area

The integration area that the X-Ray Background Tool uses to convert the average map values (which are given in units of 10-6 PSPC counts s-1 arcmin-2) to count rates with units of PSPC counts s-1 is the number of map pixels used to calculate the average map value multiplied by the dimensions of each pixel in arcmin2. The pixels in the ROSAT All-Sky Survey diffuse background maps have dimensions of 0.200 degrees by 0.200 degrees or 144 arcmin2.

Converting the Output Count Rates to Fluxes Using WebPIMMS

The HEASARC Web Tool WebPIMMS, a flux/count rate converter, can be used to convert the average map values (which are given in units of 10-6 PSPC counts s-1 arcmin-2) to flux values (in units of ergs s-1 cm-2 arcmin-2) via the following recipe:

  1. Calculate the integration area, A, over which the X-ray background count rate has been calculated, in arcmin2. A suitable integration area based on the number of pixels used to calculate the average map values is provided on the output page of the X-ray Background Tool.
  2. Convert the average map value(s) of the X-Ray Background Tool into PSPC counts s-1 by multiplying by 10-6 * A, where A is the integration area calculated in step 1.
    Note: You can skip steps 1 and 2 if you check the Integrate Count Rates Over Search Area option in the input form of the X-Ray Background Tool. The tool will then perform the necessary calculations and present the average count rates with units of PSPC counts/sec in the output page.
  3. Go to the WebPIMMS page. In the "Convert From" box, select ROSAT PSPC using the pull-down menu. In the "Into" box, select FLUX from the pull-down menu.
  4. Enter the appropriate energy range in the "Input Energy Range" and "Output Energy Range" boxes, based on the table above. For example, if you were interested in the R1 band, you would type in "0.11-0.284" (no quotes) in each box.
  5. In the "Source: Flux/Count Rate" box, enter in the count rate (in units of PSPC counts s-1) calculated in step 2.
  6. In the "Value of nH" box, enter the an appropriate NH value; a reasonable value would be the NH value returned at the bottom of the X-Ray Background Tool output. For example, if the NH value returned was 1.8E+21 H-atoms cm-2, enter 1.8E+21 in the NH box on the WebPIMMS page.
  7. Now choose an appropriate emission model. Note that there are a limited number of choices which will limit the accuracy of the overall flux determination. In general the X-ray background is rather complicated and can only be accurately modeled using multiple component emission models. However, a reasonable model would be a Raymond-Smith thermal model with a temperature kT = 0.3 keV. To choose this model, select the "Raymond-Smith" button in the "Model of Source" column, and type 0.3 in the "keV" box.
  8. Press the "Estimate Count Rate" button and WebPIMMS will return the calculated value of the flux in ergs s-1 cm-2. To get the surface brightness, divide this result by the integration area A calculated in step 1.
  9. Those interested in the most accurate conversion of X-ray background count rates to fluxes will need to use WebPIMMS with multiple component models. For details of such sophisticated modeling procedures, see Kuntz and Snowden (2000, ApJ, Vol. 543, p. 195).

Converting the Output Count Rates to Fluxes Using XSPEC

A more accurate way to convert the X-ray background average count rates to fluxes is to use XSPEC, or any other OGIP-compatible spectral fitting package. Checking the option Create XSPEC-compatible FITS spectrum will result in a FITS spectrum file dynamically generated specifically for the search region requested. The cosmic spectrum and appropriate ROSAT response matrix can be downloaded from the output page. A scientifically reasonable spectral model includes three thermal components where one is unabsorbed and two are absorbed by the Galactic column density and an absorbed power law. The unabsorbed thermal component represents emission from the Local Hot Bubble (e.g., Snowden et al., 1998, ApJ, Vol. 493, p. 715) and should have a temperature near 0.1 keV. The power law component represents the contribution of unresolved extragalactic sources (primarily AGN) and has a photon index of 1.45. The two absorbed thermal components represent the low temperature halo which should have a temperature of about 0.1 keV and a hotter component with a temperature of 0.3 keV (with large possible variations). Since the spectrum has only seven data points, this model is perhaps a bit excessive. Note, however, that as the evaluation direction approaches the Galactic plane, or, if it includes significant discrete objects such as supernova remnants or clusters of galaxies, the true emission spectrum becomes even more complicated.

An XSPEC TCL script to facilitate the modeling process is available. You'll also need to download the ROSAT PSPCc response matrix. The following is the recipe for using this script:

  1. Generate the spectrum for the region desired using the X-Ray Background Tool. Download the resulting FITS spectrum, the response matrix, and the aforementioned TCL script to your current working directory.
  2. With XSPEC installed and the appropriate environment configured, start XSPEC by simply typing the command "xspec" at the command line.
  3. At the XSPEC prompt, type
         XSPEC> source xspec-sxrbg.tcl
    This assumes the script is located in your current working directory. If it is not, give the full path to the location of the script you downloaded above.
  4. At the next XSPEC prompt, type
         XSPEC> sxrbg spectrumfile pspcc_gain1_256.rsp nHvalue
    where spectrumfile is the name of the FITS spectrum that was generated by the X-Ray Background Tool and nHvalue is the value of NH, the Galactic H I column density, with units of 1022 atoms/cm2. (Some adjustment to the value of NH reported by the X-Ray Background Tool may be necessary to match those units. For example, if the tool reports an NH value of 5.80E+21 atoms/cm2, then the value that should be supplied to XSPEC is 0.580.)

Please refer to the XSPEC Users Manual for more information on how to use XSPEC.

Counts Statistics (cstat) and Counts-Based Spectra

Counts statistics are an alternative method for generating a background spectrum. Using the count rates (Ri) and count rate uncertainties (Ui) that are already calculated, an effective exposure (Ei) and number of counts (Ci) in each band are determined. For each band (i), the effective exposure is

Ei = Ri / (Ui * Ui)

and the number of counts is

Ci = (Ri * Ri) / (Ui * Ui)

The effective exposure varies because the number of background counts subtracted to make the rates and uncertainties are being ignored here.

The count-weighted average exposure (<Ew>) for the spectrum is then calculated, like so:

<Ew> = [sum(Ci * Ei)]/[sum(Ci)]

The effective number of counts for a given band i, Cei, and the statistical uncertainty in the effective number of counts, Uei, for each band are deteremined thusly:

Cei = Ri * <Ew>
Uei = Ui * <Ew>

These values are then populated into a counts-based FITS spectrum (dtype = 1 in the HEASoft wtpha2 routine).

Command-Line Tool

The source code to a command-line version of the X-Ray Background Tool is in the public domain and available for download. Power users interested in computing the average X-ray background count rates for a multitude of sky regions should consider using the command-line tool for batch jobs. The source code requires HEASoft to be installed.


  • sxrbg.tar.gz - source code (refer to enclosed README file for build instructions)
  • sxrbg-data.tar.gz - RASS data files used by the command-line tool


If you use this tool (either the command-line tool or the web interface) in your research, please cite it using the BibCode 2019ascl.soft04001S. ascl:1904.001


ROSAT Survey Diffuse Background Maps, Paper II
    S. L. Snowden, R. Egger, M. J. Freyberg, D. McCammon, P. P. Plucinsky,
    W. T. Sanders, J. H. M. M. Schmitt, J. Trümper, and W. Voges,
    ApJ, Vol. 485 (1997), pp. 125-135

First Maps of the Soft X-Ray Diffuse Background from the ROSAT XRT/PSPC All-Sky Survey
    S. L. Snowden, M. J. Freyberg, J. H. M. M. Schmitt, W. Voges, J. Trümper,
    R. J. Edgar, D. McCammon, P. P. Plucinsky, and W. T. Sanders,
    ApJ, Vol. 454 (1995), pp. 643-653

Analysis of ROSAT XRT/PSPC Observations of Extended Objects and the Diffuse Background
    S. L. Snowden, D. McCammon, D. N. Burrows, and J. A. Mendenhall,
    ApJ, Vol. 424 (1994), pp. 714-728

Deconstructing the Spectrum of the Soft X-Ray Background
    K. D. Kuntz and S. L. Snowden
    ApJ, Vol. 543 (2000), p. 195

Progress on Establishing the Spatial Distribution of Material Responsible for the 1/4 keV Soft X-Ray Diffuse Background Local and Halo Components
    S. L. Snowden, R. Egger, D. P. Finkbeiner, M. J. Freyberg, and P. P. Plucinsky.
    ApJ, Vol. 493 (1998), p. 715


Software developed and maintained by Edward J. Sabol of the HEASARC.

This software is based primarily on the research of Dr. Steve Snowden, who also assisted extensively with the algorithms used, testing, and enhancement requests.

Special thanks to Micah Johnson for his extensive assistance with XIMAGE integration.

Many thanks to Dr. Keith Arnaud for his assistance with XSPEC scripting and for his late-night guidance in writing FITS spectrum and region extensions.

Additional software testing and feedback by Dr. Stephen Drake and Dr. Mike Corcoran.

This file was last modified on Wednesday, 20-Nov-2019 22:28:51 EST

For feedback or questions concerning the X-Ray Background Tool, please contact Edward J. Sabol.

HEASARC Home | Observatories | Archive | Calibration | Software | Tools | Students/Teachers/Public