KICP Seminar: Ben Thorne (UCSC)

12:00–1:00 pm Zoom

Ben Thorne (UCSC) "A Bayesian Approach to CMB Lensing Reconstruction and Galactic Foreground Removal"

A major goal of the next decade of CMB experiments is the mapping of CMB polarization, and the search for the distinctive signature of Inflation imprinted upon it. To extract all the available information from these highly sensitive observations, we will need to improve upon the current standard quadratic estimator used in CMB lensing reconstruction, whilst also robustly accounting for polarized emission originating from our own Galaxy. In this talk I will present a framework which leverages modern Bayesian analysis and machine learning tools to jointly account for the effects of CMB lensing and Galactic foreground contamination in the search for primordial gravitational waves. I will describe a recent successful application of this framework to data from the South Pole Telescope, and the future application to observations from the South Pole Telescope and BICEP Array. Finally, I will illustrate the flexibility of the Bayesian approach by presenting on-going work to implement alternative non-Gaussian dust models. 

Event Type


May 6