Research Scientist, Climate Modeling
About this role
The Climate Modeling team at the Allen Institute for Artificial Intelligence builds advanced machine learning methodologies to improve traditional physics-based climate models, including their fast autoregressive open-source emulator, ACE. For this three-year project funded by google.org, the team is developing an open-source, AI-first climate model named NeuralGCM+ that couples Google Research's NeuralGCM with an ML-based ocean model to achieve accurate multidecadal forecasts. The research scientist will design and test strategies for coupling NeuralGCM with an ocean model, decompose physical parameterizations into physical equations and learned parts, and publish results in peer-reviewed journals while collaborating closely with external partners and mentoring postdocs.
This role exists to advance climate forecasting by building lightweight, physics-informed hybrid models that generalize effectively to unseen future climates and foster community-wide adoption. A strong candidate combines fluency in Python and machine learning with deep expertise in earth science or atmospheric physics, demonstrating a track record of peer-reviewed research and collaborative open-source development.
What you will do
- Design, implement, and test strategies for coupling NeuralGCM with an ocean model.
- Develop approaches to decompose physical parameterizations and constraints into physics-based equations and learned components.
- Present research findings at major AI and climate modeling conferences and author peer-reviewed publications.
- Engage in open-source code development through daily team meetings, code reviews, and design documents.
- Mentor postdoctoral researchers and interns joining the lab.
What you need
- Substantial experience building and deploying relevant machine learning applications in a practical or academic research setting.
- Fluency in Python and collaborative coding practices.
- Ph.D. in computer science, computational science, applied mathematics, statistics, atmospheric science, or a related geophysical science.
- At least one accepted conference paper or peer-reviewed first-authored publication utilizing machine learning.
- Demonstrated excellent communication skills.
Nice to have
- Formal training or practical experience with physically-based atmospheric, oceanic, or related model development.
- Previous professional or post-doctoral experience applying machine learning skills.
- Experience writing machine learning code in JAX.
- Formal graduate-level machine learning coursework.
Benefits
- Generous bonus plans to provide a competitive compensation package
- Generous paid vacation and sick leave
- Family leave
Skills and tech stack
Questions candidates ask
What are the remote work arrangements for this role?
Persons in these roles are expected to work from the offices in Seattle, though on-site requirements vary based on position and team.
What is the compensation range for the position?
The base salary range is $167,030 - $260,570, supplemented by generous bonus plans to provide a competitive compensation package.
How long is the project funded for?
This position is supported by a grant for a three-year project with a current period of performance ending September 22, 2029, with potential for longer-term funding if the project demonstrates exceptional progress.
What is the starting point for the new model?
The starting point is the Google Research NeuralGCM model, a hybrid ML model that uses a conventional discretization of winds encoded in JAX and learns a column-local representation of all other atmospheric and land processes from historical data.
The application is on the official The Allen Institute for Artificial Intelligence careers page.
Apply on company site