STIS CCD Imaging Throughputs

The results of on-orbit measurements of the STIS/CCD imaging throughputs are reported. Wavelength-dependent aperture corrections and revisions to detector and filter throughput curves are presented. The throughput of the F28X50LP filter is found to be close to prelaunch estimates. Initial on-orbit measurements of this filter had suggested a 28% lower throughput, but we find instead that the overall sensitivity of the STIS CCD itself needed to be decreased by as much as 18% at long wavelengths. Wavelength-dependent aperture corrections are also important for the reddest stars, with as much as 20% of the flux outside of a 0.5” radius. Introduction The Space Telescope Imaging Spectrograph (STIS) was installed in the Hubble Space Telescope (HST) in 1997, as part of the second HST Servicing Mission (SM2). During the subsequent Servicing Mission Orbital Verification period (SMOV2), the measured count rates in F28X50LP images were uniformly low by about 28% with no apparent color dependence (Ferguson et al. 1999). In response, the pre-launch F28X50LP throughput was reduced by 28% over most of its wavelength range. This reduced throughput F28X50LP curve was the version implemented in the synphot1 file stis_f28x50lp_005_syn.fits and the calstis pipeline reference file jba1538mo_apt.fits. The prelaunch 50CCD throughput curve (combined detector and mirror throughputs are tabulated in the synphot file stis_mirvis_006_syn.fits and are folded together with the OTA throughput in the 1. See Bushouse and Simon (1998) for a description of the synphot synthetic photometry package. Copyright© 1999 The Association of Universities for Research in Astronomy, Inc. All Rights Reserved. Instrument Science Report STIS 2004-05 pipeline file h7o0858mo_pht.fits) gave correct results for GD 71, a hot white dwarf standard star, but there was no appropriate data to determine the color dependence of the CCD imaging modes. Here, we discuss revisions made in 2001 to these throughputs based on additional data obtained during cycles 7, 8, 9, and 10. Relative Filter Throughputs As STIS has both spectrographic and imaging capabilities, the on-axis throughput of any of its filters as a function of wavelength can easily be checked by comparing slitless spectra taken through clear and filtered apertures. In April 1998, as part of the calibration program 7661, such slitless spectra were taken of the hot standard star BD +75 ̊325, using the STIS G230LB, G430L, and G750L gratings through the F28X50LP, F28X50OII, and F28X50OIII filters as well as through the unfiltered 50CCD aperture. CCD spectra were also taken using a number of MAMA filters to directly measure those filters’ redleaks. The F28X50OIII filter’s redleak extends to wavelengths longer than 10200 Å, where overlap of the second order spectrum prevents accurate calibration of the G750L grating; as a result of this, the long-wavelength behavior of this redleak is poorly constrained. Table 1. Slitless Spectra used for measuring relative throughputs of CCD imaging filters. To derive the throughput for each filter, the net counts observed through that filter are divided by the net counts observed for the corresponding observation through the clear Aperture Grating cenwave (Å) exposure times (s) data set name 50CCD G230LB 2375 6 o4om010s0 F28X50OII G230LB 2375 720 o4om010t0 F28X50OIII G230LB 2375 720 o4om010u0 50CCD G430L 4300 5 o4om01010 F28X50LP G430L 4300 84 o4om01090 F28X50OII G430L 4300 84 o4om010a0 F28X50OIII G430L 4300 66 o4om010b0 50CCD G750L 7751 18 o4om010d0 F28X50LP G750L 7751 60 o4om010g0 F28X50OII G750L 7751 720 o4om010h0 F28X50OIII G750L 7751 600 o4om010i0 50CCD G750L 8975 20 o4om010k0 F28X50LP G750L 8975 50 o4om010n0 F28X50OII G750L 8975 671 o4om010o0 F28X50OIII G750L 8975 720 o4om010p0