Thorium Argon is this gas which one lets through the spectrograph to get emission lines. Because Th-Ar has lots of emission lines, it is relatively easy to get a wavelength scale (which order & line has what wls) for each pixel and then use it as a wl-solution to the whole observing run. The point is that wl-scale changes from one setting (weather, spectrograph type, filters, etc.) to another and it was a year-long project to write IDL-software that produces wl-scales for the whole spectra using only 1 order identifications (user would identify about 10-15 lines by hand and give the order # and then this software package would find wls for each pixel, assuming non-uniform dispersion (that was the hard part) for each order). This document is the atlas -- the EXAMPLE Th-Ar spectra -- which one should use to identify lines in that 1 order and then use some software (ours may become public some day) to get wl-scale.
After getting wl-scale for the particular setting, one can look at stellar spectra and identify lines (get their wls) -- sharp absorption features in the star. From these it is possible to tell which element is present in the stellar photosphere, what processes are going on in the star, its redshift and other useful information. We hope to put spectra of many stars and other astronomical objects with line identifications on the web within the next 2 months.
The document contains thorium-argon spectra taken with two different spectrographs: Hamilton type and High-Resolution type.