JPL Home NASA Home Caltech Home
Follow this link to skip to the main content
NASA Logo - Jet Propulsion Laboratory
JPL Home Earth Solar System Stars & Galaxies Technology
Science Visitor and Colloquium Program
Information for Visitors
Special Cosmology Seminar

A Probabilistic Approach to Astronomical Survey Exploitation Applied to Dark Energy Constraints from Cosmic Shear
Presented by Michael Schneider
Lawrence Livermore National Laboratory

Monday, March 26, 2018
1:00 P.M. in 169-336

Upcoming wide-field optical surveys such as the Large Synoptic Survey Telescope (LSST) and WFIRST have ambitious goals for the precise measurement of cosmological and dark energy parameters. For the gravitational lensing distortion of galaxy images, or "cosmic shear", a primary systematic error is the unknown distribution of intrinsic galaxy shapes. By simultaneously inferring the distribution of galaxy morphologies disambiguated from lensing in a hierarchical probabilistic model, we achieve large gains in gravitational lensing precision and learn about the evolution of galaxies over cosmic time. Many probabilistic algorithms become computationally challenging however when working with a large number of images taken at different times or from different instruments. We have developed computationally efficient and flexible algorithms to combine inferences about image features from different epochs or telescopes using importance sampling of "interim" single-epoch inferences. I will demonstrate how our importance sampling algorithm allows us to solve previously intractable computing problems such as the marginalization of point-spread function uncertainties in individual epochs while improving characterization of sources common to all epochs. I will also show how our probabilistic forward-modeling approach enables new types of cosmological measurements with large-scale structure.

JPL Contact: If you would like to meet with the speaker, contact Jason Rhodes (4-3304).

SVCP Astrophysics

Privacy / Copyrights
  NASA Home Page
Site Manager:

CL 08-3220