Published:
We used generative diffusion models to sharpen coarse climate data over the U.S., showing they can recover fine-scale weather features and provide useful uncertainty estimates for risk analysis.
Published:
Machine learning parameterizations are becoming a popular technique for improving climate models. This work aims to learn the uncertainties associated with them.
Published:
Work led by Rob King to compare calibration and uncertainty quantification techniques for a gravity wave parameterization
Published:
Visualizations of Calibrate, Emulate and Sample method used in Mansfield & Sheshadri, 2022.
Published:
A video of the emissions - response emulator built during my PhD.