How AI is Making Climate Modeling Faster, Greener, and More Accurate

Christopher Bretherton, Senior Director of Climate Modeling at the Allen Institute for AI (AI2), highlights how AI is revolutionizing climate science. In this NVIDIA GTC 2024 session, Bretherton presents advancements in machine learning-based emulators for predicting regional climate changes and precipitation extremes. These tools accelerate climate modeling, making it faster, more efficient, and accessible for local and regional planning.

By bridging the gap between traditional physics-based models and the need for cost-effective, high-resolution forecasts, AI-powered simulations are empowering researchers and policymakers to tackle climate challenges with greater precision. They also enhance extreme weather planning while significantly reducing the environmental footprint of climate simulations.

Key takeaways include:

AI2 Climate Emulator (ACE): Developed by AI2 using Spectral Fourier Neural Operator (SFNO) architecture, ACE accelerates climate simulations by 1000x while reducing power consumption by 10,000x compared to traditional models. Training on 100 years of NOAA model data takes just 2.5 days on four NVIDIA A100 Tensor Core GPUs, and a 100-year simulation runs in three hours on a single A100. Enhanced realism and skill: ACE reproduces extreme rainfall patterns and climate variability with accuracy matching state-of-the-art DOE and NOAA models, supporting robust long-term forecasting. Generative ML for downscaling: Techniques like video super-resolution enhance spatial resolution, delivering detailed precipitation predictions crucial for planning in different regions.

Watch the session Machine Learning Has Taken Weather Forecasting by Storm. How About Climate Modeling? Explore more videos on NVIDIA On-Demand, and gain valuable skills and insights from industry experts by joining the NVIDIA Developer Program.

This content was partially crafted with the assistance of generative AI and LLMs. It underwent careful review and was edited by the NVIDIA Technical Blog team to ensure precision, accuracy, and quality.