Postdoctoral Research Associate - Improving Sea Ice and Coupled Climate Models with Machine Learning

Princeton University seeks a Postdoctoral Research Associate to develop machine learning-enhanced sea ice models in collaboration with NOAA's GFDL. The role involves creating a neural network for model corrections, implementing it in the SIS2 sea ice model, and evaluating its impact on climate simulations. Candidates should have a Ph.D. and relevant experience in sea ice physics, coupled climate modeling, or machine learning. The position requires in-person work at Princeton University.
The Atmospheric and Oceanic Sciences Program at Princeton University, in association with NOAA's Geophysical Fluid Dynamics Laboratory (GFDL), seeks a postdoctoral or more senior research scientist to develop hybrid models for sea ice that combine coupled climate models and machine learning. Our previous work has demonstrated that neural networks can skillfully predict sea ice data assimilation increments, which represent structural model errors (https://doi.org/10.1029/2023MS003757). When applied online to global ice-ocean simulations, this neural network substantially improves sea ice simulation performance (https://doi.org/10.1029/2023GL106776). The successful applicant for this position will work to develop a conservative machine-learning based sea ice model correction that can be applied to fully coupled climate model simulations. The project will involve: 1) the development of a neural network that conserves heat, mass, and salt across model components; 2) implementation of the network in the SIS2 sea ice model; 3) evaluation of the impact of the conservative network in fully coupled historical and scenario climate simulations; and 4) the development of an emulator for sea ice model physics. The work is part of a larger project, M2LInES, covering eleven institutions. The overall goal is to reduce climate model biases at the air-sea/ice interface by improving subgrid physics in the ocean, sea ice and atmosphere components of existing coarse resolution IPCC-class climate models, and their coupling, using machine learning. The postdoc will be expected to collaborate with other postdocs at Princeton and with other members of the M2LInES project across multiple institutions. In addition to a quantitative background, the selected candidates will ideally have one or more of the following attributes: a) familiarity with concepts of sea ice physics and polar climate; b) experience with coupled climate modeling; and c) experience with machine learning. Candidates must have a Ph.D. or expect to complete a Ph.D. for an anticipated start date by summer 2025. The Term of appointment is based on rank. Positions at the postdoctoral rank are for one year with the possibility of renewal pending satisfactory performance and continued funding; those hired at more senior ranks may have multi-year appointments. Complete applications, including a cover letter, CV, publication list, research statement (no more than 2 pages incl. references), and 3 letters of recommendation should be submitted by February 28, 2025, 11:59 pm EST for full consideration. Princeton is interested in candidates who, through their research, will contribute to the diversity and excellence of the academic community. Applicants should apply online at %listings_link%. For additional information contact Dr. Mitch Bushuk (mitchell.bushuk@noaa.gov) or Dr. Alistair Adcroft (aadcroft@princeton.edu). The work location for this position is in-person on campus at Princeton University. This position is subject to Princeton University's background check policy which will include meeting the security requirements for accessing the NOAA Geophysical Fluid Dynamics Laboratory. Princeton University is an Equal Opportunity Employer and all qualified applicants will receive consideration for employment without regard to age, race, color, religion, sex, sexual orientation, gender identity or expression, national origin, disability status, protected veteran status, or any other characteristic protected by law.