The Spitzer/IRS ATLAS project

• Home
• History
• Project
• Data
• Contrib

a a a a a a a a a a a a a a a a a a a a a a a a a a

The origin of ATLAS

by Antonio Hernán Caballero

It all started back in 2006, while I was working on my PhD Thesis. I had a sample of 70 Spitzer/IRS spectra of intermediate and high redshift sources (the ELAIS-IRS sample) and wanted to compare the properties of these galaxies with the many other samples of low and high redshift sources that back then where starting to appear in the literature.

The problem was that published measurements of key spectral properties, such as the equivalent width of PAH features, were not directly comparable, because every author used his own methods (including myself). So I needed to get to their spectra and measure the properties with my own tools. Oh, boy! easier said than done: many of the spectra were still in their proprietary period, so they couldn't be retrieved from the Spitzer Archive, and the owners wouldn't easily give their highly valuable data to a competing astronomer. On the other hand, the fully reduced and calibrated spectra were already available, plotted in the figures of published papers. If only there were a way to extract these data...

My PhD advisor, Dr. Ismael Pérez Fournon, showed me this way in fact existed, since preprints in the Arxiv archive can be downloaded as source TeX and postscript figures, and the postscript file is in essence a set of instructions to the postscript interpreter on how to draw the figure, including the vector data, i.e., the points of the spectrum. So I did a little of research on how the information is encoded in the postscript format, found the sequence of postscript commands responsible for plotting the spectrum, and wrote a python script to extract them and convert values from postscript coordinates to physical units. The script worked so well that I could build a collection of almost 200 spectra in a few weeks, while the reduction of my own 70-strong sample had taken the best part of a year.

After my thesis was finished in 2008 and the usefulness of the collection demonstrated, I decided to expand the project in the form of an Atlas of MIR spectra of star-forming and active galaxies (the ATLAS project) and make it available to other astronomers. Soon my friend Evanthia Hatziminaoglou joined with a bunch of ideas on how to exploit the data, and also looked for new papers with more spectra to include in the sample and obtained multiwavelength photometry from several VO archives.

By mid-2010 we had more than one thousand spectra from around 750 different galaxies, with consistent measurements of MIR spectral properties as well as optical, near-, mid- and far-infrared and radio photometry for most of them. This is the ATLAS sample we are now releasing to the wider community.