Page History

...

Q3: I completed background subtraction in the pipeline using median background subtraction and it seems to consistently under subtract the background (and never over subtract), especially on the red end.  Is there another option I can try, or am I doing this incorrectly?  

A3: One potential source of the problem is that there are known ~5-10% spatial throughput variations that are not yet included in the MOS calibrations (see the NIRSpec MOS Calibration status page for more details).  If the background slitlets are preferentially located on one part of the MSA plane (or preferentially fall on either NRS1 or NRS2), the median sky background may not be representative of the background in the source slitlets.  It would be good to investigate the spatial variation in the background and see if this problem persists even when using background shutters that lie near the science shutters in the MSA.

It would also be good to first determine that this is not a flux conversion problem.  A good way to probe the issue is to systematically turn off flux calibration/flat_fielding/etc.  This way you are only looking at the raw extracted spectrum and you can also experiment with each step (trying one at a time) to see what correction they are doing.  The steps to turn off to try this are: barshadow, pathloss, flat_field, photosphotom, master_background_mos.

Additionally you can look at the sky background to see if there is an intrinsic spatial variation as this may also be coming into play, particularly looking to see if you it changes as you get further away from the cluster.    

Try these options and if you need further help please send a ticket to the Help Desk and we can work more with you. 

...