A Novel Approach to Constraining Baryonic Feedback: Coherent Field-Level Inference of Lensing and Electron Pressure

One key challenge in constraining cosmology using galaxy weak lensing measurements is developing an accurate model of how gas is distributed in the universe. The mismodeling of the gas distribution could potentially explain one of the outstanding issues seen in cosmology today. This research team will apply a novel framework based on numerical simulations to the latest data from cosmic microwave background (CMB) and galaxy surveys to gain insights on this topic.

This work will bridge a gap between the technical development that has taken off due to the recent advances in artificial intelligence (AI) and machine learning (ML) tools, and the deep knowledge that has been gained over the last 10-years in working with large galaxy and CMB survey data sets. This award will additionally support development of an AI/ML curriculum as well as new collaborations with the Adler Planetarium on visualizations of the universe.

Accurately modeling how baryons are distributed in the universe is challenging. The mismodeling of baryonic distribution could potentially explain the S8 tension between galaxy weak lensing surveys and CMB experiments. The investigators will apply a novel field-level inference framework based on differentiable simulations to the latest data from CMB and galaxy surveys to gain insights on this topic.

They envision a three-step process building up to a final analysis using galaxy weak lensing data from the Dark Energy Survey (DES) and the thermal Sunyaev-Zeldovich (tSZ) effect map from the deepest 100 sq. degree data from the South Pole Telescope (SPT-3G). First, the team will build on their prototype pipeline and scale up the volume and resolution of the weak lensing portion to what would be required to analyze the data.

Second, the team will implement the tSZ effect into their field-level simulations. Finally, they will incorporate further observational effects into the simulations and apply the framework to the DES and SPT-3G datasets.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.