As already mentioned in the December 2010 STAN, one consequence of the gain sag in the COS FUV detector is the mis-registration of events along the cross-dispersion or y-direction, commonly referred to as y-walk. While this y-walk has no effect on the extracted spectra, it could potentially affect FUV dispersed-light target acquisitions (PeakXD step). This kind of Target Acquisition uses information about where the target's spectrum is found in the cross-dispersion direction vs. where it is supposed to be found (determined from the location of the wavecal) to be properly centered in the aperture. Targets that are initially well centered in the aperture will be more affected by y-walk than targets that are not well centered initially, as the latter will fall on a more pristine area of the detector where the gain has not sagged as much. We estimate that as of early 2011, COS FUV dispersed light target acquisitions could lead to target mis-centerings of the order of ~0.4"-0.5" with G160M, ~0.35" with G140L, and 0.3" with G130M and this effect will get worse as the gain sags even more.
Analysis of data obtained during SMOV showed that the target's flux starts to be compromised when the mis-centering is ~0.5" or larger.