[Previous] | [Session 14] | [Next]
R. Nichol (Carnegie Mellon Univ.), Computational AstroStatistics Team, SDSS Collaboration
Clusters of galaxies are key tracers of the mass in the universe. Therefore, by studying their distribution and evolution, we can obtain strong constraints on the underlying cosmological model. This is especially true once cluster observations are combined with CMB and Supernovae data to form the ``Cosmic Triangle'' (Bahcall et al. 2000). To obtain high precision cluster observations, we need new, larger, catalogs of clusters that span a wide dynamic range in both redshift and mass as well as possessing a well-understood selection function. In this talk, I will outline the construction of such a catalog through the combination of the Sloan Digital Sky Survey and the ROSAT data (both archival pointings and the all-sky survey). This catalog is predicted to contain of the order of 10,000 clusters/groups out to z~q0.7 and thus provide a tight constraint on \Omegam, the mass density of the universe (\Delta\Omegam < 10% at the 3\sigma level). To facilitate in the construction of this new catalog, we are employing new efficient statistics developed as part of a new collaboration - called ``Computational AstroStatistics'' - between Astrophysicists, Computer Scientists and Statisticians. I will review these new methodologies which include the use of the EM algorthm in classifying X-ray sources, the use of Mixture Models to simultaneously cluster X-ray and optical data and the use of multi-resolutional KD-trees for fast correlations between catalogues. This research is partially funded via an NSF KDI grant.