2MASS Northern Observatory Survey Camera

J Band Bias Slope Problem - Software Fix

The Northern Survey Camera has developed an electronics problem which manifests as bias level slopes, discontinuous at the quadrant boundaries, which vary from frame to frame in J Band. This problem causes background discontinuities on the order of the background noise level in the frames and coadded images, degrading galaxy and point source detection sensitivity and reliability.

A software fix has been developed which effectively eliminates the problem and minimally affects real sky structure. The fix takes advantage of the fact that the error may be approximated by a function of quadrant line number, and is correlated across the quadrants of the chip. Frexas was modified to compute the lower quartile for each line of each frame and write the values out in a file. This file is read by a small program (tiltfix) which computes the band scan coordinates for each line using the *.fpos files, registers the data spatially, and computes a robust offset and slope for each frame to minimize the residuals in the registered quartiles. The slopes and offsets are in turn read by pixphot and applied to the frame data prior to coadding. This procedure leaves in the sky structure.

This image shows the J-band line medians for each frame plotted as a function of in-scan position for 980114n/s001 before and after the per frame slope correction, and also the residuals after correction. This scan had unusually bad bias slope problems.

This image shows the J-band line medians for each frame plotted as a function of in-scan position for 971205n/s026 before and after the per frame slope correction, and also the residuals after correction. This scan passed directly over m33, and shows negligible systematic effects due to the large true slopes in background.