The COS/FUV detector is subject to "gain sag", i.e, with usage it becomes less efficient at converting incoming photons to electrons. When a large enough area of the detector becomes affected by gain sag, it is necessary to move the spectral location to a fresh part of the detector. The first such move occurred on 23 July 2012, when the COS FUV spectrum was moved to the COS FUV Lifetime Position 2 (LP2).
At the time that the COS Cycle 22 Call for Proposals was released, it had been anticipated that the large majority of Cycle 22 COS FUV observations would be performed at the COS FUV Lifetime Position 3 (LP3), and the expected instrument performance, including a 10-15 % decrease in spectral resolution below values obtained at LP2, was documented in the Instrument Handbook and HST Primer. It now appears that the move to COS/FUV Lifetime Position 3 will not occur until early 2015, and so a substantial fraction of Cycle 22 observationsare likely to occur at the COS FUV lifetime position 2, prior to the change.
In most cases, observations done at LP2 should be adequate to achieve any science goals that were designed to be consistent with the expected Cycle 22 COS FUV performance as documented in the COS IHB and the HST Primer. However, in some cases it may be undesirable to split a program's observations between the two lifetime positions. Both the line-spread function and the profile perpendicular to the dispersion will differ significantly between the two positions, as will the details of the fixed pattern noise. There may also be small differences in throughput between the two positions. So programs that have especially stringent requirements for the repeatability of the instrument performance across visits might be best deferred until after the move to LP3 if their execution cannot be completed prior to that change.
If you believe your Cycle 22 program might be adversely affected by being split across lifetime positions, please explain in detail why that is the case in the "additional comments" section of your phase 2 proposal. Programs affected by the split will be scheduled either at LP2 or LP3, depending mostly on schedulability.
The LP3 position will be located at -2.5" below the original LP1 position. This is closer than the LP2 position which was located at +3.5" above LP1. As a result extended targets observed at LP3 may have more overlap with the gain-sagged regions near LP1 than would observations done at LP2. Previous Cycle programs that require the larger spacing at LP2 may be affected if their execution is deferred until after the move to LP3. If you have an earlier Cycle program where you are concerned that your program might be adversely affected by being split across lifetime positions or if you have any further questions, you should either discuss this issue with your assigned Contact Scientist or send your concerns to help@stsci.edu