Rémi Lam of Google DeepMind is in Nature’s Top 10. Lam, whose visionary work connects AI innovation with the challenges of our changing climate, presented the challenges and the outcomes of his research in one of the CMCC Lectures. Watch the video and find out more about how AI enhances weather forecasts.

Rémi Lam, a researcher at Google DeepMind, has been named one of Nature’s Top 10 scientists of 2024, a recognition of his groundbreaking contributions to the advancement of weather forecasting through artificial intelligence and machine learning.

Lam is reshaping how we predict the future—of weather and beyond. His pioneering model, GraphCast, uses machine learning to deliver faster, more accurate global weather forecasts, setting a new standard for the field. The model, presented in Science in November 2023, demonstrated superior accuracy compared to traditional methods, excelling in extreme event predictions such as tropical cyclones and extreme temperatures.

GraphCast, an AI neural network designed as a global grid and trained on real atmospheric data without relying on explicit physical laws, surpassed conventional physics-based weather models in accuracy, achieving remarkable results more quickly than expected.

Earlier this year, Lam joined the CMCC Foundation as the inaugural speaker in its CMCC Lectures webinar series, presenting his innovative approach to medium-range weather prediction. His work exemplifies a meeting point between cutting-edge AI and climate science, aligning with CMCC’s mission to advance innovative research for global resilience.

Extreme weather events, such as hurricanes, and innovative approaches to understanding and predicting their impacts were also the focus of the second webinar in the CMCC Lectures Series, presented by Gabriel Vecchi, director of the High Meadows Environmental Institute at Princeton University. The recording of the webinar “Linking past and future hurricane activity changes” will be available on CMCC’s YouTube channel soon.

More information:

AI weatherman: the DeepMind researcher making faster, more accurate forecasts