|
Abstract:
|
In 2014, the Large Synoptic Space Telescope (LSST) will begin mapping the entire night sky every 3 days, producing over 40 tbs of data/day. One of the major goals will be to systematically detect and identify transient changes in the sky, such as supernovae and unknown near-Earth asteroids. The LSST will need to do the near-real time analysis of very large data sets under changing atmospheric conditions, raising a host of statistical and computational problems. In this paper, we introduce a new method for reliably detecting transient changes in LSST images. Our method offers an improvement over methods currently being considered by the LSST by being much more general and relying on fewer assumptions, while still being able to be computed quickly. We explore the benefits of our method over the current method through simulation and using a data set provided by the Sloan Digital Sky Survey.
|
