How to analyse SPI data of GPS and GCDE

last update of this page: 24.05.2006

Introduction
Step-by-step guidelines
Data, lists of DOLs, and results for GPS and GCDE

Here I describe how I do the analysis of the Galactic Plane Scan (GPS) and Galactic Centre Deep Exposure (GCDE) data. You might have better ideas how to do it - then please send me your comments.
Some advice before we go into the details (if you are not familiar with SPI processing, use the SPI online user manual (very short), or follow the SPI cookbook)

Do not forget to create a logfile (e.g. by applying setenv COMMONLOGFILE myprocess.log); this can be very useful in case you run into problems.
Then select your data set which you want to analyse wisely. In case there are science windows which do not contain useable data (e.g. CONF mode, slews, science windows with zero GTI) do not include them in your list. Use IOSM to check the science windows, or use a prepared list from the core program SPI analysis page
The more data you combine, the better. But the areas should overlap.

After og_create use spi_science_analysis for the processing up to the IMA level (including SPIROS). spi_science_analysis calls all necessary executables. It also helps you not to screw up the parameter files.
Find below some advice for the single steps for usage of OSA 5.1

Step-by-Step

step important parameters remarks

General Parameters detectors="0-18" (rev. 19-139)
detectors="0,1,3-18" (rev. 140-214)
detectors="0,1,3-16,18" (after rev. 214)
use the pseudo detectors for single+PSD events. Take into account the failure of detector #2 in revolution 141, and #17 in rev. 214

General Parameters coordinates="GALACTIC" for GPS, GCDE, GLS use Galactic coordinates.

CAT_I catalog="spi_bright_cat_0002.fits[1]"
fluxMin="0.0001" use the input catalog spi_bright_cat_0002.fits as your general reference catalog and use a Minimum flux of 0.0001

BIN_I nregions = 1
regions="20,40"
nbins="1" standard binning used in analysis

BKG_I nmodel="1"
model01="GEDSAT" When applying SPIROS background method "5", the background model produced here is not used.

IMA
spiros background-method="5"
mode="IMAGING"
image-fov="POINTING+ZCFOV"
nofsource="3"
srclocbins="FIRST"
pointing-subset="AUTO" this is for IMAGING. Run it once, and check for pointings with high reduced chi squared (at the end of the logfile). Exclude them from the analysis using the pointing-subset parameter in spiros. Check the logfile for sources which turned out to have significance 0.0 sigma. Remove these sources from the source_cat.fits file by setting their SEL_FLAG (last column) to 0
Then re-run the spi_science_analysis again, but only the IMA step (un-check the CAT_I, POIN, BIN_I, BKG_I step box)
Check the output again; iterate when necessary.

cat2ds9 catDOL="source_res.fits"
fileName="regionfile.reg" Produce a region file

ds9 load spiros_image_sigma_result.fits into ds9 and use the region file you produced in the previous step