New Developments for the Cosmic Origins Spectrograph

Rachel Osten,, for the COS team

Those of us who are committed spectroscopists know the unparalleled value of a spectrum. The results of the Cycle 18 proposal selection process affirmed that many Hubble Space Telescope users feel similarly: in Cycle 18, orbits with the Cosmic Origins Spectrograph (COS) account for 23% of GO prime orbits allocated. The overwhelming majority of these orbits use COS far-ultraviolet (FUV) spectroscopic configurations. COS/FUV spectroscopy also accounts for about 29% of Snapshot (SNAP) orbits.

The decline in sensitivity of the COS FUV spectroscopic configurations continues, but the behavior has changed. Discovered midway through Cycle 17, the phenomenon was described in the instrument science report (ISR), COS 2010-15, “Early Results from the COS Spectroscopic Sensitivity Monitoring Programs,” by Osten et al.

The monitoring program uses repeated observations of spectrophotometric white-dwarf standard stars to measure any time dependence to the spectroscopic sensitivity. The main finding in the ISR for the FUV detectors was that the decline increased for longer wavelengths, down by about 4–6% per year at short wavelengths and by 12% per year at the longest wavelengths. Observations taken since mid March 2010 have exhibited a flattening of the slope at the longest wavelengths. As a result, the FUV data are now consistent with a value almost independent of wavelength: a weighted mean of –5.1 ± 0.2%/year (see Figure 1). This result suggests that the sensitivity decline of the FUV detector has two origins: one is producing a more or less constant level of sensitivity decline, and another—now past—causing the initial decrease. The updated time-dependent sensitivity reference file reflects this new information, with a change in the FUV slope occurring on the date 2010.2.

The COS team at the Institute is setting up routine monitors of the data quality, enabling automated checking of the instrument’s performance. Examples include monitoring the accuracy of acquisitions; automated monitoring of the dark current on both the FUV and near-ultraviolet (NUV) detectors; routine searches for short time-scale, high-voltage transients; weekly images of the accumulated photons falling on the detector; and monitoring the decline in the brightness of the lamp. The results of these monitors will be available from the main COS web pages.