101. LSST Data Management: Building the Data System for the Era of Petascale Optical Astronomy
- Author
-
Juric, Mario
- Subjects
LSST ,Astronomy ,Optical Surveys ,Data Management - Abstract
The Large Synoptic Survey Telescope (LSST; http://lsst.org) is aplanned, large-aperture, wide-field, ground-based telescope thatwill survey half the sky every few nights in six optical bands from320 to 1050 nm. It will explore a wide range of astrophysicalquestions, ranging from discovering “killer” asteroids, to examiningthe nature of dark energy. The LSST will produce on average 15 terabytes of data per night,yielding an (uncompressed) data set of over 100 petabytes at theend of its 10-year mission. To enable the wide variety of plannedscience, the LSST Project is leading the construction of a new,general-purpose, high-performance, scalable, well documented,open source data processing software stack for O/IR surveys.Prototypes of this stack are already capable of processing datafrom existing cameras (e.g., SDSS, DECam, MegaCam), and formthe basis of the Hyper Supreme-Cam (HSC) Survey data reductionpipeline. In the 2020-ies, running on dedicated HPC facilities, thissystem will enable us to process the LSST data stream in near realtime, with full-dataset reprocessings on annual scale. In this talk, I will review the science goals and the technical designof the LSST, focusing on the data management system, itsarchitecture, the software stack, and the products it will generate. Iwill discuss the exciting opportunities it presents for LSST, and theastronomical software community as a whole. More broadly, I willalso discuss implications of petascale data sets for astronomy inthe 2020s, and ways in which the astronomical community canprepare to make the best use of them.  
- Published
- 2015
- Full Text
- View/download PDF