Columbia University, Postdoctoral Research Scientist in Machine Learning of Earth System Model Data
The Department of Applied Physics and Applied Mathematics (APAM) at Columbia University invites applicants for a Postdoctoral Research Scientist position in machine learning and climate science at its Morningside campus in New York City.
The Postdoctoral Research Scientist will focus on using machine learning (ML) to develop improved and flexible ML emulators of diverse Earth System Model (ESM) inputs and outputs. The emulators will serve as inexpensive-to-run, accurate surrogates for ESM components. Bayesian methods will be applied to the emulators (with outputs scored against existing aggregated observations) so as to efficiently optimize ESM parameter settings, thus improving ESMs and enabling more skillful climate projections. The emulators and parameter inference code will be trained and tested on output from GISS ModelE (atmosphere) as well as output from models at the National Center for Atmospheric Research (NCAR; both from the CESM atmosphere [CAM] and land model [CLM]). The Postdoctoral Research Scientist will work closely with Dr. Gregory Elsaesser (APAM & NASA GISS), Dr. Marcus van Lier-Walqui (Center for Climate Systems Research [CCSR] & NASA GISS), and a supportive team that includes machine learning experts and atmospheric scientists at both the National Center for Atmospheric Research (NCAR) and the Center for Learning the Earth with Artificial Intelligence and Physics (or LEAP, an NSF Science and Technology Center (STC) at Columbia University launched in 2021; https://leap.columbia.edu).
The Postdoctoral Research Scientist will be expected to organize and chair a LEAP research working group related to ML – parameter inference in models, lead peer-reviewed publications, and to present their results at LEAP and other scientific meetings. The Postdoctoral Research Scientist should also expect to occasionally (though infrequently) travel to NCAR as part of multi-institutional collaborations.
- A Ph.D. in Data Science, Bayesian Statistics, Computer Science, Physics, Earth System Science, Atmospheric Science, or a directly related discipline is required by the start of the appointment.
- Strong programming skills are a requirement.
- Fluency in Python.
- Advanced experience in machine learning.
- Demonstrated experience in statistical/mathematical analyses of model output and/or observational datasets.
- Excellent command of the English language (verbal and written) and strong communication skills are desired.