7. WHAT IS THE PHOTOMETRIC PRECISION OF THE FOS CALIBRATION

The overall uncertainty in the absolute FOS flux calibration is estimated at ~3%, which is a ~2 s value derived from the combination of a 1-2% uncertainty in the model WD flux slopes, a fraction of a percent uncertainty in the Landolt photometry used to normalize the models to absolute flux, and a 1-2% internal FOS photometric repeatability (Bohlin 1995). The internal consistency is demonstrated to be better than ~2% by Figure 5, which shows the residuals of the sum of all FOS cycle 4 observations in comparison to the standard stars on the WD scale. For a case where the model flux distributions are used as the standard, the high S/N co-added spectrum for G191B2B has residuals of <2% in high dispersion on all wavelength scale-lengths, except in the spectral lines, where small wavelength errors and resolution mismatches can cause bigger problems. The H78 region is somewhat worse because of lower S/N and more uncertainty in the flats. The low dispersion modes are less photometric, especially for the prisms and at the Balmer lines, where the resolution is badly mismatched. Actual FOS observations of point sources may have more uncertainty than the 3% associated with the calibration files, especially if the pointing uncertainty exceeds 0.1-0.2 arcsec.

[Previous] [Next] [Up]

Generated with CERN WebMaker