[Previous] | [Session 12] | [Next]
J. Kim-Quijano, I. Busko, J.A. Valenti (STScI), D. Lindler (ACC, Inc.), C. Bowers (GSFC)
We describe the performance of a new two-dimensional algorithm for correcting scattered light in echelle spectra obtained with the Space Telescope Imaging Spectrograph (STIS). The algorithm was developed by D. Lindler (ACC) and C. Bowers (NASA/GSFC) and is described by these authors in a separate contribution (2000AAS...1971202) at this meeting. STScI has implemented this new algorithm in the publically available IRAF package, CALSTIS, which contains tasks to reduce data from the four cross-dispersed echelles in STIS.
One-dimensional interpolation of the background between echelle orders does not adequately describe the background beneath STIS echelle orders, leading to errors in residual intensity as large as 10% in the far-UV. The new algorithm iteratively builds a two-dimensional model of the scattered light based on an extracted spectrum and known properties of the spectrograph.
We present examples of spectra extracted with both the one-dimensional and two-dimensional algorithms for removing scattered light. We then use the zero-intensity cores of saturated interstellar absorption lines to characterize the accuracy of both algorithms as a function of wavelength and echelle grating. Finally, we discuss how to retrieve and run the new package, as well as plans to use the algorithm when serving reduced data from the Hubble archive via the On-The-Fly-Calibration (OTFC) system at STScI.