Towards Efficient and Precise Queries over Ten Million Asteroid Trajectory Models

Abstract

The new generation of telescopes under construction return to the same area of the sky with sufficient frequency to enable tracking of moving objects such as asteroids, near-earth objects, and comets [4,5]. To detect these moving objects, one image may be subtracted from another (separated by several days or weeks) to differentiate variable and moving sources from the dense background of stars and galaxies. Moving sources may then be identified by querying against a database of expected positions of known asteroids. At a high-level, this task maps onto executing the query: “Return all known asteroids that are expected to be located within a given region at a given time.” We consider the problem of querying for asteroids in a specified interval in space and time, specifically as applied to populating the simulations of the data flow from the Large Synoptic Survey Telescope (LSST).