Straylight Correction in IRS SL

An illustration of the average straylight correction (SLC) in a source with intermediate zodiacal foreground (NGC2798):

Straylight Black=no change, Red=SLC added flux, Blue=SLC removed flux

This map was produced by averaging 36 pointings in a spectral map, then differencing the BCD and F2AP (SLC-free) versions. The PU fields at right have been scaled and replicated from the averaged BCD frame. The map color is coded with black near zero difference, red as positive difference (straylight correction added flux: up to 25% change in SL1), and blue as negative. Typical positive horizontal striping is easily seen, but rolling substructure is also visible. The correction in SL1 (at left) should be very nearly zero, but it does not decay rapidly away from the peak-up fields (not visible at right). The broad negative correction at 12-13um is visible at left.

An example of various extractions using either the BCD or F2AP (from Kartik Sheth):

NGC2798 Extractions

The large and spurous negative adjustments to both spectral orders at the point of spectral overlap (7.6um) conspires to cause them to meet there. The spurious bump at 8.6um blends directly into a real PAH feature there. That this bump is mostly spurious can be seen from examining the spectral cube at 8.57um:

NGC2798 Cube Slice

At this wavelength over half the light in an aperture which covers the sub-slit length is contributed by the horizontal bands from straylight mis-correction.

A less pathological case is for NGC4826: here the aggregate SLC is much smoother (although it does still add flux):

Straylight Black=no change, Red=SLC added flux, Blue=SLC removed flux


Update on S11 REMOVSTRAYCROSS module

Jim Ingalls re-ran the REMOVSTRAYCROSS module for the droop product, to yield the following two differences (this time rsc - droop), both scaled in the same way:

Black=no change, Red=SLC added flux, Blue=SLC removed flux: S10.5a

Black=no change, Red=SLC added flux, Blue=SLC removed flux: S11

Note the absence of the order imprint, which was being injected by the flat-field. A cut across the array, totaled over rows 76:78 illustrates the exponential fall-off, and the non-zero pedestal flux:


Last modified: Tue Sep 28 10:26:41 MST 2004