Visual Discovery in Large-Scale Astrophysical Datasets; Experiences Using the Sloan Digital Sky Survey

Abstract

Nowadays astronomers are experiencing an unprecedented growth in the quality and quantity of datasets coming from numerical simulations and real-world observations. For example, the increasing availability of high performance computing facilities has given the possibility to perform large-scale simulations of several dimensions. Also, forthcoming astronomical surveys are expected to collect datasets of several petabytes. The emerging need is thus for efficient visual discovery tools for rapid inspection to identify regions of interest in large-scale datasets prior to applying computationally expensive data analysis algorithms. This paper reports our experiences in developing visual discovery tools for the Sloan Digital Sky Survey (SDSS), the most ambitious astronomical survey ever undertaken. We present existing tools and visualization requirements collected from SDSS users for new functionality. We then discuss a range of newly developed visual discovery tools and their applicability to SDSS and finally we conclude with pointers to future developments.