The COS FUV detector is a windowless cross delay line device based on a photon-counting micro channel plate design with two independently operable segments (FUVA and FUVB). Each segment has a distinct dark current that may be correlated to the Solar Cycle (see Figure 1 and Figure 2). The dark rate, in particular for FUVA, has been decreasing recently, which in concert with improved algorithms in CALCOS to filter out background counts as a function of pulse height leads to substantially smaller dark rates. The ETC value for the dark rate is currently the same for both segments and was based on a conservative estimate of an FUVA dark rate of 4.37x10-6 cnts/s/pixel (the dark rate for FUVB is lower by a factor of ~2). This was the correct value at the time of the ETC update. However, the dark rate on both detectors has been evolving with time, particularly in the last few months, and it is now a factor of two lower than the value currently used in the ETC for both segments. In a majority of programs, the source count rate will be larger than the dark current, but PIs considering observations of faint, background limited targets will need to adjust their ETC results.
For most modes in the FUV, one can recalculate the signal-to-noise ratio (SNR) and exposure times directly (i.e., see: http://etc.stsci.edu/etcstatic/users_guide/1_2_time_snr.html). Observers will want to note the source and the dark counts (the first and fourth values of the second column of ETC output in the detailed information section) to use the equations listed in the link above. In the case where the source count rate is equal to the dark current, one would multiply the expected SNR by a factor 1.15, and the expected exposure time by a factor of 0.75.
For the new blue modes (the 1055 and 1096 central wavelengths of the G130M grating), users should consult: http://www.stsci.edu/hst/cos/software/planning/etc/ETC_blue_modes.pdf. The ETC does not accurately reflect recent improvements to the focus and resolution of these modes and requires users to calculate corrections to the default ETC output. The corrections depend on the ratio of source count rate (C) to dark count rate (Bdark), α = C/ Bdark. For these corrections, users should multiply their calculated α from these prescriptions by a factor of two to account for the current lower dark rate.
Investigators are warned that STScI cannot guarantee that the dark rate will be lower than the current ETC value when their observations execute. If the impact to the number of requested orbits is minimal, PIs might want to consider using the conservative dark rate currently implemented in the ETC. Assuming values lower than 3.0x10-6 cnts/s/pixel for FUVA and 2.0x10-6 cnts/s/pixel for FUVB when calculating the time allocation needed for the Phase I of proposals will be at the observer’s own risk.
The Cycle 21 Phase II ETC will be updated with segment-dependent dark rates and will take into account changes to the new blue modes due to the focus change (resolution and extraction height).