The MINT group works with a range of external collaborators to pioneer the next generation of generative machine learning system that will inform the science, technology, and policy in mitigating and adapting to a changing climate. Our work broadly centers on three pillars at the intersection of artificial intelligence and climate change.

  • Research: We regularly publish papers at the frontiers of artificial intelligence and climate change. Our current focus areas are in physical sciences. See representative works papers here and here.

  • Engineering: We develop and maintain dataset benchmarks and software for accelerating research and engineering efforts.

  • Education and Outreach: We develop accessible courses, tutorials and seminar talks to onboard students, researchers, and practitioners with different backgrounds. We organize events that engage diverse stakeholders.

news

  • ChaosBench accepted for oral presentation at NeurIPS 2024 – top 1.8% of the papers. Congrats Juan!

  • Stormer accepted for publication at NeurIPS 2024.

  • Stormer won the best paper award at the Climate Change in AI (CCAI) Workshop at ICLR 2024. Congrats Tung and Hritik!

  • We released Stormer, which beats Graphcast to set a new state-of-the-art for ML-driven weather forecasting.

  • Thank you Washington Post, MIT Tech Review for covering our research on ClimaX. Aditya’s comments on the field more broadly appeared in Nature and Science.

  • Aditya gave keynote talks on Stormer, ClimaX, and ClimateLearn at Hydro ML Conference, KDD Environment Day, and UCSD Scientific ML Symposium, seminar talks at Columbia, Cornell, Argonne National Lab, UT Austin, UC Berkeley, University of Michigan, and University of Washington.

  • ClimateLearn v1 and benchmark preprint released!

  • Tung invited to give a seminar talk on ClimaX at Leeds SciML Seminar.

  • ClimaX also wins best paper award at ICML Workshop on Synergy of Scientific and Machine Learning Modeling which will take place right after the main conference. Congrats Tung!

  • ClimaX accepted for publication at ICML main conference. See you in Hawaii!

  • Tung and Aditya, along with collaborators from Microsoft, recently released ClimaX, the first foundation model for weather and climate. We are excited by the public and press so far, and have many interesting followups in store.

  • We launched our PyTorch library ClimateLearn and humbled by the public response so far.

  • Aditya is chairing the AI4ClimateScience Bridge Program at AAAI on Feb 8, 2022.

  • Hritik, Tung, Shashank, and Aditya presented a tutorial on ML for predicting Climate Extremes at NeurIPS 2022 workshop on Tackling Climate Change with Machine Learning. Selected for spotlight!