Petascale Object Classification of the LSST Event Stream

Details Petascale Object Classification of the LSST Event Stream Petascale Object Classification of the LSST Event Stream

Jan 2009

adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2009aas...21346029b&link_type=abstract

American Astronomical Society, AAS Meeting #213, #460.29; Bulletin of the American Astronomical Society, Vol. 41, p.372

Other

1

Scientific paper

The LSST object database will contain detailed information for 20 billion sources, including approximately 10 billion galaxies and a similar number of stars. After 10 years of LSST survey operations, the object database will comprise 10-20 Petabytes of science catalog attributes: over 200 science attributes per object will be available for classification, characterization, and mining. Deep co-added values (in multiple passbands) for static objects and long-term time series (at various cadences) for dynamic objects, based upon about 1000 individual observations of each, will yield an enormously rich potential for new scientific discoveries. As part of this petascale discovery process, the impressive quantity and quality of parameter data for each object will enable the classification of astronomical objects on a grand scale. This is especially critical for the LSST event stream -- the LSST is likely to detect 10 to 100 thousand astronomical events per night. An event is defined to be any source that has changed in position and/or brightness relative to the baseline "template sky". In order for the astronomical research community to assimilate, cope with, and process such an enormous nightly flood of events, it is essential to develop and deploy a petascale object classification pipeline. This science pipeline will generate object classifications and likelihoods, based upon spatial data (e.g., positional coincidences and associations), temporal data (e.g., LSST photometric and astrometric time series), and VO-accessible data (i.e., corroborating catalog data within other data repositories at this sky location). These classifications will permit knowledge-based prioritization of the most significant events for follow-up time-critical observation. We describe some specific examples using ANN (Artificial Neural Networks) based on data from SDSS (Sloan Digital Sky Survey).

Affiliated with

Also associated with

No associations

Rating

Petascale Object Classification of the LSST Event Stream does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Petascale Object Classification of the LSST Event Stream, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Petascale Object Classification of the LSST Event Stream will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-1706179