Open Postdoctoral Position in Machine Learning and Climate Science in Mandt/Smyth Groups

January 2025

We have an open position at UC Irvine for a postdoctoral research fellow to work at the intersection of machine learning methods and climate science, to be jointly advised by Stephan Mandt and Padhraic Smyth.

This postdoctoral position will build on existing collaborations and prior work in the Mandt and Smyth groups at the interface of machine learning and climate science, involving (for example) heavy-tailed diffusion models, generative models for precipitation data, uncertainty quantification in spatio-temporal forecasting, and more. Current climate science collaborators include Mike Pritchard (UCI/NVIDIA), Jim Randerson (UCI), Efi Foufoula-Georgiou (UCI), and Pierre Gentine (Columbia University).

Qualified applicants should apply at: UCI's online computer science application portal. Applications will be reviewed on a rolling basis until the position is filled. The position is funded for two years.

Requirements

• PhD in a quantitative discipline (computer science, statistics, mathematics, engineering, physics, etc)
• Record of publications at leading machine learning venues such as NeurIPS, ICML, AI-Statistics, etc
• Interest and/or experience in both (a) developing new machine learning methods and (b) addressing important problems related to climate and the environment

Expectations

The successful candidate will be expected to

• Publish at leading machine learning conferences
• Participate in scientific collaborations and publish in climate/environmental science journals
• Lead projects involving PhD students in the Mandt and Smyth groups

Recent Relevant Papers from Mandt and Smyth Groups

For examples of recent work from our groups at the interface of ML and climate science, see:

• Heavy-tailed diffusion models, Pandey et al, arXiv 2024
• Precipitation downscaling with spatiotemporal video diffusion, Srivastava et al, NeurIPS 2024
• A generative diffusion model for probabilistic ensembles of precipitation maps conditioned on multisensor satellite observations, Guilloteau et al, arXiv 2024
• Functional flow matching, Kerrigan et al, AI and Statistics 2024
• ClimSim: A large multi-scale dataset for hybrid physics-ML climate emulation Yu et al, NeurIPS 2023
• Climate-driven changes in the predictability of seasonal precipitation, Li et al, Nature Communications 2023
• Comparing storm resolving models and climates via unsupervised machine learning, Mooer et al, Scientific Reports, 2023
• California wildfire spread derived using VIIRS satellite observations and an object-based tracking system, Chen et al, Scientific Data, 2022
• Zonally contrasting shifts of the tropical rain belt in response to climate change, Mamalakis et al, Nature Climate Change, 2021