Project Info
The distribution of observed galaxies contains plenty of information about the constituents and origin of the universe. This is is traditionally extracted by approximating the probability of finding galaxies to be Gaussian, and analyzing compressed observables that are only sensitive to this Gaussian part. However, since the true galaxy distribution is not Gaussian, this compression is not lossless. In my work I explore methods to include the additional information in the non-Gaussian parts of the galaxy distribution. A novel method recently proposed with Tobias Baldauf and Uros Seljak relies on correlating fields that are quadratic in the original density with the density itself. We show that this is nearly optimal for specific applications like measuring the amplitude of fluctuations. At the same time the method is computationally extremely fast, and as simple as the estimation of Gaussian observables (power spectra). While analogous methods are well established for analyzing the Cosmic Microwave Background, this approach has not been applied to large-scale structure before.