GLASTSpec Help

GLASTSpec provides a facility for simulating spectra for different types of observations by the Fermi detectors using one of a number of spectral models. It utilizes the X-ray spectral fitting package, XSPEC. Note that this tool allows you to simulate the spectrum from only one detector at a time. If you would like to simulate joint fits, you can download the necessary FITS files (an option available after performing a spectral fit with GLASTSpec) for multi-detector fits in XSPEC.

Detector/Observation Type and Model Choice Page

On the first page, you choose a detector/observation type combination and the desired model to be used in the simulation. Both fields provide a scrolling list; more choices are available than are shown initially.

Once you have specified the detector/observation type and the desired model, click on the "Specify Model Parameters" button to continue the process.

Model Parameter Entry Page

The next page is a dynamic page. Its contents are determined by your choices on the previous page. The first thing on this page is a double check for you to see what GLASTSpec thinks you have asked it to do. Seen here are an English description of the model chosen and the detector/observation type combination chosen. Also shown is the exact expression that XSPEC is going to use as the model. If these aren't what you wanted, you should go back to the previous page and try making a different selection.

Note that all energies are in keV.

Next, you get the first (and the only blank) entry to make, the exposure. Enter the desired exposure in seconds. Now you get to enter the energy boundaries. There are two types of energy boundaries. The first set is a lower and upper boundary for the purposes of spectral fitting. This specifies the energy range over which the spectra is being fit. The second type is a way to specify energy band(s) for flux calculations. Note, that some detector/observation type combinations give the option to enter more than one band. The lower and upper energy ranges given here do NOT affect the fitting; they merely give an energy range for which a flux should be calculated.

After the energy ranges are entered, you are given one or more sets of model parameters to specify. There is a set for each component chosen on the previous page. For each parameter, the default value is already entered. For each parameter, you get the opportunity to freeze (fix at the specified value) or thaw the parameter for fitting purposes. Also, you can specify whether or not an error for that parameter should be computed. The error calculation is based on finding the parameter values for which the delta fit statistic is 2.706 from the best fit value. This is equivalent to the 90% confidence region for a single interesting parameter. Again, the default settings for each parameter are loaded when the page is first loaded.

After the exposure, energy ranges, and model parameters have been entered, click on the "Show me the spectrum!" button to submit the simulation. It may take a minute or two to finish. Once it is done, the results page will be loaded.

Results Page

If you have gotten to this page, you probably have entered things properly. The first thing offered is an image of the resulting spectrum. You can click on the image to get a larger version for closer examination. At this point, you can change the plotting specifications, (x and y scales, labels, etc.) and replot the simulated data and model. You are also offered the opportunity to download a postscript version of the plot. After the image, the model expression is echoed again, as well as the mission/instrument combination and exposure. The resulting model parameters, along with their errors where requested, are listed next. The reduced chi^2 for the model and the count rate over the fitted energy range is given next.

Next comes the flux table. A table listing the fluxes for the energy band(s) specified is displayed. This table shows the lower and upper energies for the flux calculations, the count rate, photon flux, and energy flux. Inspect these resulting fluxes. If they are not what you wanted or expected, you may wish to go back to the parameter entry page and modify the normalization(s) used. In the simplest case, you would merely multiply the entered normalization by the ratio (flux wanted)/(flux returned).

Finally, you are offered the opportunity to download or view the various files associated with the simulation. You can download the spectra (source and background) and the response matrix used, if you would like to do your own spectral work. You may also download the qdp and pco files that were used to create the plot of the simulated spectrum. Last, but perhaps most useful, you can view the XSPEC script file and the resulting log file. These can be supremely useful if something just doesn't quite look right. The log file will show you the output at every step. The script file will show you what XSPEC was told to do.


For feedback or questions on WebSpec, please contact Stephan Fantasia. fantasia@milkyway.gsfc.nasa.gov