Thanks to an innovating 3-mirror design (modified Paul-Baker), the LSST telescope will provide aberration free wide field images (9.6 deg²). The optical beam will be collected by a giant 3.2 billion pixels camera covering the focal plane (0.64 m diameter).
In 2023, LSST will start a systematical survey of the southern hemisphere covering 18,000 deg². Every area of the sky will be visited more than 800 times during the 10 years of operation of the instrument.
After co-addition, the images will allow to reach a magnitude exceeding 27.5 (100 billions times less luminous than the Vega star). The outcome of the main survey will feed a catalog of 10 billion galaxies and as much stars from our own Milky-way.
In order to guarantee that every single area of the sky will be revisited with a maximum delay of 3 days, LSST will rely on a wide field optical system and on short (2 x 15 s) exposures. The redshift of galaxies will be estimated thanks to the highly accurate photometry provided by 6 pass-band filters covering from the near ultraviolet (320 nm) to the near infrared (1320 nm).
Every night, 15 TB of raw data will be acquired. At the end of the survey, the total amount of data (raw and secondary) will reach 0.5 Exabytes and the last catalog stored in the database will index 7000 billion sources representing a total of 15 PB. Such a volume as well as the complexity of the indexing require a very sophisticated software stack running on an optimized hardware architecture. From this point of view, LSST also constitutes a challenge in the big data area.
During the survey, the data management system will generates alerts on transient events (supernovae, asteroids,...) within less than 60 seconds after the image acquisition. Catalogs of astronomical objects will be produced yearly from the collection of images acquired since the beginning of the project. These catalogs will be distributed worlwide by a database system able to handle the most sophisticated queries on a system massively distributed over thousands of computing nodes.