Toward the end of the IUE mission, two IUE Project published two reports documenting changes in the response of the SWP camera. The first of these reports, by Garhart & Turnrose (1998), concerned the functional form of the degradation of target fluxes in exposed low-dispersion spectra of IUE calibration stars. Garhart & Turnrose found that the secular decrease in SWP-camera sensitivity worsened noticeably at late epochs, and recommended that a bilinear fit be made to this degradation with a break point at 1991-2. Second, Gonzales-Riestra (1998) compared flux ratios from pairs of exposures of calibration standards taken at the same date but at different exposure levels. By dividing the measured ratios by the reciprocal of the exposure times, she derived the nonlinearity of the instrument response as a function of optimal exposure level. Her data show that these departures are small except at short wavelengths and low illuminations.5 This result hints that the sensitivity variation are caused mainly by a change in zeropoint (background surface) of the flux calibration and is thus consistent with the results in this study.
Although changes in the background surface characteristics are primarily reponsible for nonlinearities in flux at low exposure levels, they are not the only problem associated with aging of the SWP camera. For example, we note that on grayscale renditions one can discern at least two more echelle orders shortward of m = 113 in early-epoch images than in late-epoch images. This loss in sensitivity, combined with the decrease in the BCKGRD-determined background level (compared to the true zero-flux level at the Lyman core), causes a spurious increase in NEWSIPS of the absolutely-calibrated far-UV flux level, even though the sensitivity of the instrument has decreased in this region (Nichols & Linsky 1998). We note parenthetically that IUESIPS-processed spectra do not show this artifact.