The characterization of the combined effect of a non-uniform illumination pattern and the difference in illumination between sky and the internal, undispersed stimulator flashes can be calibrated by imaging celestial sources of uniform surface brightness (see Gordon et al. 2005). We have utilized SED observations of diffuse Galactic emission (with IRAS Iν(60 µm)/ Iν(100 µm) ~ 0.2-0.3) near the Galactic plane. Once an adequate number of independent observations have been acquired, a median filtering is performed to filter out any spatial structure that might be present in individual observations. To reach a good signal level (i.e., ~5-10 times the level of the dark current), we targeted regions where IRAS Iν (60µm) = 200-300 MJy sr-1 and where no IRAS point sources are identified within a radius of about 5-10 arcminutes.
Table 4.3: Column-wise Correction Factors to IC.
The current IC approach is quite efficient, reaching good signal levels over the entire detector array using a reasonable amount of telescope time. However, we found some residual systematic variation of up to ~15%, mainly as a function of the detector column. The cause for this residual is not quite clear at this point. Table 4.3 shows this column-wise IC residual derived from 16x1 raster maps of a few bright point sources with the raster step size matching the detector pixel size. Dividing Table 4.3 into the observed IC image in a column-wise way gives our composite IC that reduces the residual flux variation to within about 5% from detector column to column. This composite IC was implemented in the pipeline starting in pipeline S15.