You are here: IRS>Smart Web>CassisHome (13 Dec 2016, DonBarry)

Latest news

(LR=low-resolution atlas, HR=high-resolution atlas)

  • CASSIS now stands for "Combined Atlas of Sources with Spitzer IRS spectra"
  • URL to CASSIS:
  • 2015 June release of LR7 and HR1
  • 2015 January test of LR7 pipeline
  • 2014 June: test of HR1 pipeline
  • 2014 February 26 release of LR6 atlas version
    • 2013 November: test of LR6 pipeline
    • 2013 July 18: LR5.2 atlas version update
    • 2013 May 25: LR5.1 atlas version update
    • 2013 March 26: release of LR5 atlas version
    • 2012 April: processing of LR5 atlas version
    • 2011 November: validation of LR5 pipeline
    • 2011 mid-October: testing of LR5 pipeline. Major improvements: better image and spectra combination, bug fixes for plots, ...
    • 2011 August 15th: the CASSIS website is online! The first public version is referred to as "version 4"
    • 2011 July 20th: the CASSIS paper is accepted
    • 2011 March 8th: the CASSIS paper is submitted
    • 2011 January: processing of LR4 atlas version
    • 2010 December: testing and validation of LR4 pipeline

Spectra are available via the online database where they can be downloaded along with various other products:

--- CASSIS online database ---

The CASSIS atlas is presented in two publications. We ask users to refer to the relevant paper (or the two papers if appropriate) in their studies:

LOW-RESOLUTION "Cornell Atlas of Spitzer/IRS Sources", Lebouteiller, V., Barry, D.J., Spoon, H.W.W., Bernard-Salas, Sloan, G.C., J., Houck, J.R., Weedman, D.W., ApJS, 196, 8. (ADS services)

HIGH-RESOLUTION "CASSIS: The Cornell Atlas of Spitzer/IRS Sources. II. High-resolution Observations", Lebouteiller, V., Barry, D.J., Goes, C., Sloan, G.C., Spoon, H.W.W., Weedman, D.W., Bernard-Salas, J., Houck, J.R., ApJS, 218, 21. (ADS services)

The full citations guidelines can be found on the CASSIS website.

What is CASSIS?

Approximately 90% of the IRS observations were performed in the staring mode which observes the spectrum of a given source at two nod positions located approximately at 1/3 and 2/3 along the slit. For a given observation (called AORkey), the staring mode observes either a single source or multiple sources in a successive manner, the latter mode being also referred to as ”cluster observation”. The remaining observations include spectral mappings (7%), and moving objects (3%). Tables 1 and 2 summarizes the observations done in low-resolution (R~60-127) and high-resolution (R~600).



The Spitzer Science Center (SSC) pipeline operates in two parts, the Basic Calibrated Data (BCD) pipeline and the post-BCD pipeline. The BCD pipeline reduces the raw detector images from the individual exposures and removes the electronic and optical artifacts, including dark current, droop effect, non-linearity, radhit, jail-bar pattern, and stray light. The BCD pipeline also includes a flat-field correction. The steps performed by the BCD pipeline are relatively well understood and we consider the BCD products as reliable inputs for spectral extraction. We refer to the IRS Instrument Handbook for more details on the BCD products.


The post-BCD processing consists in combining and cleaning the exposures, removing the background, and extracting the spectrum. The post-BCD processing often requires a manual intervention or an algorithm specifically designed for a given observation. The CASSIS pipeline provides an alternative to the SSC post-BCD pipeline by enabling a flexible, automatic, extraction algorithm that is able to handle very different observations, from barely detected sources to bright sources, from point-like sources to extended sources.

The Combined Atlas of Sources with Spitzer IRS Spectra ("CASSIS") gathers the spectra automatically extracted of every source intended (as opposed to serendipitous) for observation in the staring mode of the Infrared Spectrograph (IRS) onboard the Spitzer Space Telescope. Fixed and moving objects, observed as a single source or as a cluster, are considered.

Help & Feedback

We encourage CASSIS users to join the CASSIS Users mailing list and share experiences, data recipes, advice, and problems. The CASSIS principals monitor this list and will assist as time permits. CASSIS is a work in progress! Help us make it better.

Specific and reproducible bugs can be reported through the MANTIS bugtracker (make sure to select the CASSIS project with the top/right box in MANTIS)

Low-resolution atlas (CASSIS/LR)

Version history

  • LR7 (current version)
    • A local background subtraction is performed for the nod-subtracted images, similarly to what was done for the order-subtracted images in v6 (see below). This improves significantly the agreement between the 2 individual nod spectra for very faint sources and as a result, provides a much better determination of the source's flux level.
    • For the tapered column extraction of nod-subtracted images, the spectrum can be uncertain is the source is extended since the window in which the flux is integrated could include part of the "negative" spatial peak (due to the nod image difference). In v7, the spatial extent threshold above which the use of tapered column extraction for nod-sutracted images is not reliable anymore has been fine tuned from 1.75 times the PSF width to 1.5 times the PSF FWHM.
  • LR6
    • A local residual background emission could remain for order-subtracted images in LR5. This sometimes led to contamination, especially for faint sources. The local background is now removed for order-subtracted images. The same local background is subtracted for the tapered column extraction and optimal extraction. In the case of tapered column extraction though, the local background is not removed if the source is very extended (since in this case the "local background" is likely originating from the source itself).
    • Better determination of source spatial extent. Also, when the detection level is too low and no accurate extent can be determined, optimal extraction is now chosen as the default spectrum.
    • Increased the threshold for the dispersion of coordinates over the exposures above which individual exposure images cannot be combined anymore and have to be extracted separately instead.
    • Fixed a bug when multiple spectra for a given module/order/nod have to be combined (which happens only when the images could not be combined because of a significant dispersion of the pointing over the exposures)
    • The default spectrum is now trimmed (overlaps between spectral orders are removed)
    • RA_SM and DEC_SM header keywords now give the extracted coordinates for the merged spectra
    • LR5 uses the final S18.18.0 calibration.
      • LR5.1 update: fixing bug causing uncertainties to be overestimated when 2 image exposures are combined, fixing bug on rms error not being propagated in the final tapered spectrum, the total error (column 3) for optimal spectra is now the sum of rms errors and systematic errors (version 5.0 spectra had the calibration errors added)
      • LR5.2 update: forcing order 3 to be extracted at the same position as order 2, fixing and adding some header keywords. Minor bug fixes.
    • LR4 was the first public version. It corresponds to the 4th internal iteration of the pipeline. The CASSIS LR4 pipeline uses the S18.7.0 calibration.
    • LR1, LR2, and LR3 were test versions meant only for internal use by the ISC group.

Extraction pipeline

CASSIS/LR pipeline step by step. This page explains the many steps involved in the extraction, including the diagnostics about the spatial extent of the source. Although this is not required to understand and use the spectra, we strongly encourage users to read through this page. A better use of the data will be achieved through a full knowledge of how the post-BCD CASSIS pipeline works.

Other extraction methods for CASSIS/LR. Some sources are flagged as not being a point-like source. In those cases, the tapered column extraction, or the full slit extraction should be used instead. This page explains the extraction methods. Note that as of version LR5, CASSIS chooses automatically the best extraction method for the main spectrum to be displayed and downloaded in the main result page.

Error propagation. This page explains how error are propagated through the pipeline (systematic and RMS).

Staring Cluster observations in CASSIS/LR. Please read this page to check how cluster observations are treated by the pipeline. Cluster observations are made of several pointings corresponding to one AORkey, and they require special conventions for downloading and analyzing. You will most likely not need to read this section unless you specifically worry about how cluster observations were handled.

Available products

CASSIS/LR products. What products are available, and how to use them.

Redshifts. How redshifts are found by CASSIS.

High-resolution atlas (CASSIS/HR)

Version history

  • HR1: current version

Extraction pipeline

CASSIS/HR pipeline step by step. This page explains the many steps involved in the extraction, including the diagnostics about the spatial extent of the source. Although this is not required to understand and use the spectra, we strongly encourage users to read through this page. A better use of the data will be achieved through a full knowledge of how the post-BCD CASSIS pipeline works.

Available products

CASSIS HR products. What products are available, and how to use them.

How to use the web interface

CASSIS web interface. Quick help on the interface and how to use it.

CASSIS and the Spitzer Heritage Archive

The 2 archives have different objectives and can even be used together. We list below the main specifics of each archive.

CASSIS provides:

  • the best signal-to-noise ratio available, as enabled by the optimal extraction which uses a super-sampled point-spread function,
  • fringe correction,
  • automatic choice of background subtraction method,
  • quantitative measurement of the source spatial extent and alternative extraction methods for partially-extended sources,
  • automatic selection of the best extraction method depending on the source spatial extent
  • automatic redshift determination using NED source matching,
  • a user-friendly web interface that facilitates browsing and inspection of spectra.
  • possibility to perform massive data analysis on the entire database at once

On the other hand, the Spitzer Heritage Archive provides:

  • access to mapping data (although not reduced)
  • cross-search between Spitzer instruments

Some statistics

  • The atlas currently contains about 16000 sources.
  • About 2800 sources have a redshift determination (using NED for source matching).
  • Most sources (70%) have an inferred extent below 2" and can be considered as point-like.

Publication list

Topic attachments
I Attachment Action Size Date Who Comment
pngpng cassis_newlogo_full.png manage 29.2 K 12 Dec 2016 - 10:17 VianneyLebouteiller  
pngpng tabirs.png manage 29.6 K 05 Jun 2015 - 17:50 VianneyLebouteiller  
pngpng tabobs.png manage 9.3 K 05 Jun 2015 - 17:47 VianneyLebouteiller  
pngpng tabobs2.png manage 137.6 K 05 Jun 2015 - 17:47 VianneyLebouteiller  

This site is powered by FoswikiCopyright © 2002-2018 by the IRS Team.
Send feedback. We use Foswiki.