Skysurveys, Light Curves and Statistical Challenges

Research in astronomy is undergoing a profound transformation from the study of small samples to the analysis of large-scale digital surveys of the sky. Ever larger amounts of data and better statistical techniques are being used to address a vast range of astronomical problems, from near-by asteroids to universe-wide cosmology. There is a huge need for new methodology development to address petabyte-sized datasets of images, atlases with millions of spectra, multivariate catalogues and time series with billions of objects. The focus of this article is on the analysis of light-curves (i.e. the variation of source brightness as a function of time). A description of the importance of the problem and the techniques already being used is given. The field is ripe with many statistical and computational challenges.