Statistical Methodology's Special Issue on Astrostatistics

Special Issue editorial – by guest editors:
Jean-Luc Starck, Antoine Llebaria and Tom Loredo

Modern astronomy is producing data sets of unprecedented complexity and size. The statistical and computational challenges astronomers face in distilling sound science from such data make interdisciplinary research that combines astronomy, statistics, signal processing, and computer science more important than ever. The resulting body of research constitutes a new, emerging discipline: astrostatistics. This special issue presents a diverse selection of examples of recent research in astrostatistics.

Rich and complex astrostatistical problems arise in nearly every astronomical subdiscipline, in analyses of data sets both large and small. But the main impetus for increased interest in astrostatistics has been the large data sets produced by sky surveys. Some such surveys scan large swaths of the sky; prime examples are satellite observations of the Cosmic Microwave Background (CMB), which survey the entire sky in high frequency radio waves, or ground-based optical surveys of an entire celestial hemisphere, mapping the distributions of stars and galaxies. Other surveys focus attention on a small region, but probe it with greater detail or thoroughness than is possible in an all-sky survey. Examples include “pencil beam” searches for dim, asteroid-like trans-Neptunian objects in the outer solar system, the Hubble Space Telescope’s Hubble Deep Field search for distant galaxies, or ground-based observations of small patches of the CMB, probing angular scales not accessible to satellite missions. Each data set poses unique problems for astronomers, including detecting and measuring sources hidden by complex noise and selection effects, cross-identifying sources with other data sets, classifying sources, etc. These objectives drive development of ever more sophisticated and precise statistical methods. The papers presented here span many of these problems.

Continue reading…