Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/documents/dhb/web/c07_focerror.fm8.html
Дата изменения: Tue Nov 18 00:20:13 1997
Дата индексирования: Sat Dec 22 14:57:10 2007
Кодировка:

[Top] [Prev] [Next] [Bottom]

7.8 Errors in Absolute Photometry (f/96)

The absolute photometric accuracy of FOC observations depends on several factors. This section will not discuss those sources of error that arise from errors in the flatfield correction and associated effects (e.g., pattern noise). The remaining errors most likely arise from: 1) errors in the published fluxes or variations in fluxes of the spectrophotometric stars used to calibrate the absolute DQE, 2) errors in the assumed PSFs, 3) errors in the assumed filter transmission curves, 4) format dependence effects, 5) temporal variability in the FOC detectors, and 6) the spectrum of the source. This section will summarize the current understanding (or lack thereof) of these errors. As the f/48 detector is much more poorly calibrated, it will be discussed separately. For a summary of FOC accuracies, see Figure 8.4.

Errors in the spectrophotometric standards. The spectrophotometric standards used for the FOC DQE determination are on the flux scale derived from correcting IUE spectra of the white dwarf G191B2B to conform to the pure hydrogen model of Finley (see Colina and Bohlin, AJ 108, 1931 (1994)). The spectra of the standards used here (BPM16274 and HZ4) were corrected using the same function. While it is difficult to assign a formal uncertainty to the predicted filtered fluxes due to errors in the spectrophotometry, assigning an error of +/- 3% is probably conservative enough.
Errors in the assumptions for the PSF. Because the in-orbit calibrations relied on large aperture photometry, there should be very little sensitivity to details of the PSF or changes in the PSF. This source of error should contribute less than 1% error to the derived efficiencies. (Note that quite the opposite is true when deriving total fluxes of stars from core-aperture or PSF-fitting photometry techniques).
Errors in the assumed filter transmission curves. Although the filter transmission curves were carefully measured on the ground, that does not preclude some sort of subsequent degradation or change in performance. There has been no unambiguous evidence for changes in any particular filter's bandpass. There is some evidence that the redleaks of some filters differ significantly from their published values.
Format dependence. A variation of sensitivity with video format has been noted. In particular, Table 7.3 shows the relative response of the more common f/96 formats with respect to the 512 x 512 imaging format. These determinations are not known completely accurately. Most of the absolute sensitivity calibration observations used the 256 x 256 format, so the uncertainty in the calibration of the format dependent sensitivity for this format enters into the uncertainty for all the formats. The uncertainty is approximately 3%. No such table has been derived for f/48. Note that if the image is calibrated using the PHOTFLAM from the image and the PHOTMODE keyword value indicates the format used, then no re-calculation of the absolute sensitivity is required.
Variability of f/96 DQE. The overall throughput of the FOC has been monitored over the three years before the first servicing mission, and in the UV since the servicing mission. The only evidence for change has been an ~3% decline in the sensitivity over three years, independent of wavelength. From the time COSTAR was installed until mid-1996, there was no significant sensitivity change in the ultraviolet, but a slow downward trend of approximately 10% per year has been seen in the UV since then.
Source spectrum. The value of PHOTFLAM averages F over the bandpass. Situations where the detected flux distribution is skewed in wavelength can lead to large errors in assigning the absolute sensitivity calibration to the adopted (pivot) wavelength, especially when the wideband filters are being used or where redleak plays a significant part. If there is any doubt as to whether there are significant color effects, observers are advised to use synphot or focsim to check their absolute fluxes. FOCSIM is an FOC simulator that can be run under IRAF at STScI or from a WWW form found on the FOC world wide web pages. This error is very dependent on the filter being used and the source spectrum, so no rules of thumb about its magnitude can be given.

[Top] [Prev] [Next] [Bottom]

stevens@stsci.edu