Previous
abstract 
Next
abstract
Session 18 - STIS.
Display session, Wednesday, January 07
Exhibit Hall,
The maximum signal-to-noise (S/N) ratio achievable with STIS for most spectral modes in the ultraviolet spectral region is set by the quality of the flat-fields and the photometric stability of the MAMA detectors (see Kaiser, et al., these proceedings). It is possible, however, to achieve high S/N spectra for bright targets, independently of the quality of the flat-field, using a special observing strategy wherein multiple spectra are obtained through up to five ``FP-SPLIT" slits. Targets observed through these slits project to positions on the detector that are displaced from one another along the dispersion axis (only). This observing technique permits a simultaneous solution for the target spectrum and the local flat-field granularity of the detector, in a similar manner to that of Lambert, et al. (1993, ApJ, 420, 756), who derived very high S/N spectra from GHRS data.
The standard star BD+28\hbox^\circ4211 was observed for about eight orbits with STIS and each of the gratings E140M and E230M. Several spectra were obtained with each FP-SPLIT slit, and the exposure times were kept short (typically 360 to 600 s) to avoid significant Doppler smearing within single integrations. We selected for analysis one of the best exposed spectral orders from each grating in which there were a small number of relatively narrow spectral features. The resulting S/N, as measured over a line-free segment of the continuum, was 274 for E140M, and 262 for E230M, using this technique. This compares to the Poisson limits of 337 and 283, respectively, based on the realized count levels. It is important to point out that the results obtained here are in many respects special, and will not be typical of GO science observations. Nevertheless, the derived S/N of \approx 300 is well in excess of the pre-flight goal of S/N > 100:1, and shows that very high quality spectra can indeed be obtained with STIS.
