Ph.D. Candidate: Sowmya Kamath
Research Advisor: Patricia Burchat
Date: Friday, February 21, 2020
Time: 1:30 PM
Location: Physics and Astrophysics Building, Room 102/103
Title: CHALLENGES FOR DARK ENERGY SCIENCE: COLOR GRADIENTS AND BLENDED OBJECTS
Some of the most exciting scientific challenges at present are in understanding the nature of dark matter and dark energy. The billions of galaxies observed by Legacy Survey of Space and Time (LSST) will dramatically improve the statistical power of weak lensing observations and help probe the mass distribution in the Universe. This increased statistical sensitivity means that potential systematic biases must be carefully identified, quantified, and minimized. My talk will address two such systematic biases: galaxy color gradients and blending.
In the first part of the talk, I will describe how shape measurements of galaxies with varying spectral energy distribution across their profile -- called “color gradients” -- when seen by a chromatic point spread function can be biased. I estimate this bias expected for the LSST using simulations of parametric galaxies and realistic galaxy images. The second half of my talk focuses on the blending challenge for the LSST where a significant fraction of the lensed galaxy images will overlap with images of other objects, affecting the accuracy of flux and shape measurements. I discuss two novel approaches to infer the presence of objects that go undetected because of blending and compare their performance to existing detection algorithms.