domeckert
commented
2 years ago Thanks for reporting this issue @KovacsOrsi .
I indeed think that the spike you are seeing comes from regions of very low exposure. One quick way of checking it is to set a low threshold on the exposure, and ignore the pixels that have a very low effective exposure. The exposure map can be accessed as the "exposure" attribute of the Data class. You can modify it in the following way:
expo = np.copy(dat.exposure) # get the exposure map
maxexp =np.max(expo) # compute the maximum exposure
lowexp = np.where(expo < 0.1 * maxexp) # here we cut out regions where the exposure is less than 10% of the maximum, you can adjust that
expo[lowexp] = 0.0 # regions with 0 exposure are ignored
dat.exposure = expo # replace the exposure map in the Data object
It may be worth adding an option to set automatically an exposure threshold in the surface brightness extraction, I'll consider that
I extract the surface brightness profile over a large mosaic image, which exhibits some data gap. I input a background map too (Chandra's stowed background), and its profile displays an unusual bump at a radial range where I have no full data coverage. Is this related to the the exposure/area normalization?