Data Quality Analysis at the Spitzer Science Center

Download this Presentation


Presentation Transcript

  • 1.SOFTWARE TOOLS FOR DATA QUALITY ANALYSIS The Science Data Analysis Tool (SDAT) is the primary tool used by the DQA team. SDAT can access all raw and processed data comprising any Spitzer Astronomical Observing Request (AOR). It can display images and spectra, and supports basic image analysis and histogram display. It also allows the DQA analyst to record AOR status flags and comments in the Spitzer operations database. The lower-right two images illustrate some of SDAT’s visualization capabilities, using as examples a channel-1 IRAC map of the astronomical source IRAS 05413-0104 and its channel-3 counterpart. This image is comprised of three single-color mosaics, each of which is separately examined and graded by the DQA team. They were generated by the SSC’s Mopex pipeline from Spitzer Infrared-Array-Camera (IRAC) data. The cluster, also known as the Seven Sisters and Messier 45, contains hundreds of stars, but only a handful can be seen by the unaided eye. The entire scene contains thousands of stars and galaxies.IRAC channels 2, 3 , and 4 are shown in the blue, green, and red planes, respectively. The SSC’s Mopex pipeline found 9938, 8910, 2871, and 1541 useable point sources in channels 1-4, respectively; 14,137 point sources were band-merged. Position statistics show alignment of band-merged sources of about 0.05 arcseconds. SPITZER DATA PROCESSING AND QUALITY ANALYSIS Data Quality Analysis at the Spitzer Science CenterVincent Mannings and Russ R. LaherSpitzer Science Center (SSC), California Institute of Technology, Pasadena, CA 91125 Contact: SPIE Paper No. 6270-73 ABSTRACT Data Quality Analysis (DQA) for astronomical infrared maps and spectra acquired by NASA's Spitzer Space Telescope is one of the important functions performed in routine science operations at the Spitzer Science Center (SSC) of the California Institute of Technology. A DQA software system has been implemented to display, analyze and grade Spitzer science data. This supports the project requirement that the science data be verified after calibration and before archiving and subsequent release to the astronomical community. The software has an interface for browsing the mission data and for visualizing images and spectra. It accesses supporting data in the operations database and updates the database with DQA grading information. The system has worked very well since the beginning of the Spitzer observatory's routine phase of operations, and can be regarded as a model for DQA operations in future space science missions. Keywords: data processing, data quality analysis, data archive, data mining, Spitzer Space Telescope README FILE DOWNLOADED WITH SPITZER PRODUCTSQA Summary Page Date: 2006-04-19 16:16:50.000This file provides:1. Basic information on the request (instrument, target name, etc.);2. The results of data quality analysis at the Spitzer Science Center (SSC);3. An accounting of the request's raw science data (DCEs); and4. Pipeline software-version information.(1) DESCRIPTION OF OBSERVING REQUEST (AOR or IER)Telescope: SpitzerCampaignId: 772 (IRAC006000)ReqTypeName: AORReqModeName: IracMapReqKey: 10738432TargetName: IRAS 05413-0104Program Title: H2_OUTFLOWSProgram ID: 3315RequestTitle: hh212_irac_mapObservation Start: 2005-02-22 16:30:56.202Observation End: 2005-02-22 16:47:06.207This request was a "cold" observation.(2) DATA-QUALITY-ANALYSIS RESULTSAOR quality status: Nom RlsePlease see details of the Spitzer science archive, pipeline upgrades, analysis tools, and instrument-specific data handbooks.(3) SUMMARY OF REQUEST'S SCIENCE-DATA CONTENTExpected number of DCEs: 192Received number of DCEs: 192Number of missing DCEs: 0Number of received DCEs with missing FITS-header lines: 0Number of received DCEs with missing image-data lines: 0(4) PIPELINE SOFTWARE-VERSION INFORMATIONIRAC Channel-1 Software Version: S13.2.0IRAC Channel-2 Software Version: S13.2.0IRAC Channel-3 Software Version: S13.2.0IRAC Channel-4 Software Version: S13.2.0 ASSESSMENT OF SPITZER DATA QUALITY TO DATE KEY DQA ACTIVITIES AND VALUE-ADDED A team of data quality analysts sample the data from each AOR to search for anomalies and grade the data. The data are reprocessed at the end of the campaign, and permanently stored in a publicly accessible data archive for retrieval by users worldwide. DQA status info and other meta-data are stored in the operations database. Upon retrieval by a user, the SSC’s archive retrieval tool (Leopard) triggers the construction of a text-formatted README file that includes the results of the DQA analysts’ study of the data, which is packaged with the data downloaded by the user. (See upper-right panel on this poster.) The DQA software system consists of a collection of command-line Perl scripts that are executed in the DQA Unix environment and web-based CGI scripts, Java applets, and dynamically-generated HTML documents that are executed and viewed in a web browser. The loose-integration architecture and scripting nature of the software allow software upgrades to be made very quickly. Also, scripts are, by far, the easiest kind of software to patch into Spitzer operations when bugs are found. is a Perl package that is often used at the SSC for rapid development of CGI scripts. Perl packages for accessing the Informix database are also used by the software. The necessity of people as an integral part of the DQA process was recognized early in the development of the Spitzer mission and this was incorporated into the design philosophy of the DQA software system. The first rule of data, before any kind of data analysis is applied, is to “look at the data” and humans do this best. An example CGI script from the DQA software suite SPITZER THREE-COLOR MAP OF THE PLEIADES STAR CLUSTER AOR stands for “astronomical observing request”. Three-color map courtesy of John Fowler and John Stauffer. After the DQA grading process, a status flag containing the grade is assigned to each Astronomical Observing Request (AOR) in the observing campaign. Also embedded in the status flag is information about whether the AOR will be re-observed and whether the data associated with it will be released to users.As of April 2006, the distribution of statuses for all science AORs obtained during routine Spitzer operations for all three instruments is shown at the right. Some 75 AORs have been repeated (0.39% of 19,318 science AORs). The distribution of statuses for AORs for each instrument is very similar to the overall distribution.