Enabling Massive Correlation Functions in the LSST Era
Emilio Donoso
LSST offers an exciting opportunity to characterize how galaxies cluster in the universe. A huge volume sampled at unprecedented depth and carefully calibrated photometric redshifts, will allow to accurately trace the two-point auto-correlation function, or the cross-correlation function between peculiar subsamples of objects (e.g. AGN candidates) and neighboring galaxies.
Nevertheless, the sheer size of LSST catalog/image data and the difficulties to build masks suitable for generating random sources impose new challenges to estimate even these basic cosmological tools. With this mind, we have integrated high speed pair-counting algorithms (Gundam) along with novel pixelized hierarchical catalogs working at file level (Hipscat/LSDB), to enable the distributed computation of your favorite clustering estimator for billions of point sources and for large variety of science cases. This package works tightly coupled with a framework that automatically generates pixelized masks and random samples, to provide an end-to-end solution of this problem. Here, we will explain the algorithms, technologies and code packages, as well as its application to study the projected cross-correlation between Quaia quasar candidates and over 100 million HSC-SSP galaxies, a fair example using an LSST-like massive dataset.
Event Timeslots (1)
Thu
-