A.1 COS Lifetime Positions

To mitigate the effect of gain sag, COS FUV spectra are obtained at multiple positions on the detector that are offset from each other in the cross-dispersion direction. These are known as Lifetime Positions (LPs). Details about the lifetime positions used throughout the operation of COS on-orbit can be found in Table A.1, and Figure A.1 illustrates the location of each Lifetime Position on a map showing 2D spectral profiles on both detector segments. For operations up to and including LP4, all FUV spectroscopic modes (G130M, G160M, and G140L) were moved to a new lifetime position when it was introduced. The exceptions to this statement are the blue modes (G130M/1055 and G130M/1096), which were first commissioned at LP2, and remained there until moving to LP7 at the beginning of HST Cycle 33. Starting with LP5, the decision was made to only move specific settings to new LPs, and to operate in a manner where observations at different settings would execute at different LPs. This decision was made to make more efficient use of the detector, as some settings primarily add counts to the FUVA segment while others primarily add counts to the FUVB segment, and because spectra from the different gratings have different cross-dispersion widths, such that G160M modes can make use of regions on the detector that G130M and G140L cannot. The most up-to-date description of the lifetime position at which each setting is observed can be found on the COS2025 Policies page.

Table A.1: Information about COS Lifetime Positions

The LP4 move was accompanied by a new set of restrictions on detector segment usage, FP-POS selection, and target acquisition settings. These changes, which will remain in effect at subsequent LPs, are together known as the COS2025 policy, described at http://www.stsci.edu/hst/instrumentation/cos/proposing/cos2025-policies. The main culprit of gain sag in the COS/FUV detector is geocoronal Ly α emission at 1216 Å that fills the 2.5" diameter COS science aperture whenever the G130M grating is used with the 1291/1300/1309/1318/1327 cenwaves on Segment B. In order to mitigate this gain sag, the COS2025 policy minimizes the number of locations on the detector where the geocoronal Ly α emission falls by not supporting the Segment B G130M 1300/1309/1318/1327 cenwaves and reducing the number of FP-POS for 1291 from four to two. In this way the geocoronal Ly α emission falls on fewer locations on the detector, and each LP can be used until the continuum level reaches a low modal gain. Additional restrictions aimed at reducing continuum level gain sag caused by the use of exposure times beyond those that provide the maximum achievable signal-to-noise ratio for a given observation are described in the COS2035 policy, introduced in HST Cycle 33.

Each change in Lifetime Position has resulted in a change in the spectral resolution, and small changes in the core of the line profile and the wings of the line spread function (LSF). The spectral resolution at different Lifetime Positions has been characterized (LP2: COS ISR 2013-07, LP3: COS ISR 2017-06, LP4: COS ISR 2018-07, LP5: COS ISR 2023-08, LP6: COS ISR 2023-02, LP7 and LP10: COS ISR 2026-09), and those studies find a decrease of 5–10% for LP2 and LP3 below LP1, and another 10–15% below LP3 for LP4. The resolution at LP5 is about 10% better than at LP4, and the resolution at LP6 is about 10% worse. The resolution tabulated in Table 1.1 corresponds to the current default LP for each mode for the FUV cenwaves. The resolution as a function of Lifetime Position is shown in Figure 1.1. The shapes of the line profiles are well represented in the LSF models available to the community at http://www.stsci.edu/hst/instrumentation/cos/performance/spectral-resolution.

The flux and wavelength calibrations are performed in a similar way for all Lifetime Positions. Calibration observations are executed in preparation for lifetime moves, and updates to calibration reference files are made. These observations include measurements of the spatial and spectral resolution, verification of the wavelength scales, verification of the FUV BOA operations if available, and flux and flat field calibration. In addition, LP3 and beyond make use of profile calibration files for the two-zone extraction described in Section 3.2.1, while LP1 and LP2 use a boxcar extraction (as of July 2025, two-zone extraction has been implemented for the blue modes at LP2).

The calibration pipeline automatically uses the correct calibration methods based on the Lifetime Position specified in the header keyword LIFE_ADJ, which also enables CRDS to determine the most up to date reference files for that Lifetime Position. LIFE_ADJ can have values of 0, 1, 2, 3, 4, 5, 6, 7, 10, 11, –1, –11, and –999. A value of 0 corresponds to pre-launch data. Values of 1, 2, 3, 4, 5, 6, 7, 10, and 11 correspond to those Lifetime Positions, and as new Lifetime Positions are added they will be designated 12 and higher. The values 8 and 9 were skipped. A value of –1 is specified if the aperture position is at a non-standard location. A value of –11 is only used in reference files and indicates that the file should be applied to all data with LIFE_ADJ=1 and –1. Lastly, a value of –999 is used for dark exposures, where there is no defined LP.

The varying Lifetime Positions are just for the FUV, and NUV spectroscopic observations remain at LP1 of the NUV detector.