shengchaochen82/Awesome-Foundation-Models-for-Weather-and-Climate

Curates cutting-edge foundation and task-specific models across diverse data modalities—time series, spatio-temporal video, graphs, and text—with links to papers, implementations, and datasets. Systematically organizes architectures from RNNs and Transformers to neural operators and LLM-based approaches, alongside a published survey paper evaluating foundation model adaptation for weather/climate tasks. Includes production models like GraphCast, Pangu-Weather, and Aurora alongside emerging techniques like masked autoencoders and in-context learning for meteorological forecasting.

284 stars. No commits in the last 6 months.