Back to Volume
Paper: Analyzing Data from Astronomical Surveys: Issues and Directions
Volume: 371, Statistical Challenges in Modern Astronomy IV
Page: 121
Authors: Loredo, T.J.
Abstract: Astronomical surveys arise in nearly every astronomical subdiscipline, and the resulting source or event catalogs span a huge range of sizes, from dozens to 109. Despite this enormous diversity, several common themes arise in survey data analysis, many reflecting the fundamentally hierarchical nature of survey data analysis: the intimate interplay between analysis of data for individual sources or events in surveys, and analysis of the surveyed population as a whole. Much attention has been paid to accounting for effects of truncation and censoring in suveys. I instead focus on the statistical challenges that arise when there are significant uncertainties in source characteristics (either from measurement error or “intrinsic scatter” of source properties). The issues range from calibration of luminosity and redshift indicators, to propagation of source uncertainties through population-level analysis, including measurement error, “soft” truncation and censoring, and treatment of Eddington, Lutz-Kelker, and Malmquist biases. For statisticians, I present an overview of how these issues arise in astronomy. For astronomers, I summarize some of the activity of the SAMSI Surveys and Population Studies working group, identifying areas of research in statistics with promise for resolving outstanding issues in astronomical survey analysis.
Back to Volume