FourCastNet: Accelerating Global High-Resolution Weather Forecasting using Adaptive Fourier Neural Operators

TL;DR

FourCastNet is a deep learning model that significantly accelerates global weather forecasting, achieving near state-of-the-art accuracy at a fraction of the computational cost and time of traditional physics-based methods.

Contribution

This paper introduces FourCastNet, a scalable deep learning Earth system emulator that dramatically reduces forecast computation time while maintaining high accuracy.

Findings

Predicts medium-range weather five orders-of-magnitude faster than NWP.

Achieves 140.8 petaFLOPS on supercomputers with 11.9% efficiency.

Enables high-resolution forecasts and large ensembles for better extreme weather capture.

Abstract

Extreme weather amplified by climate change is causing increasingly devastating impacts across the globe. The current use of physics-based numerical weather prediction (NWP) limits accuracy due to high computational cost and strict time-to-solution limits. We report that a data-driven deep learning Earth system emulator, FourCastNet, can predict global weather and generate medium-range forecasts five orders-of-magnitude faster than NWP while approaching state-of-the-art accuracy. FourCast-Net is optimized and scales efficiently on three supercomputing systems: Selene, Perlmutter, and JUWELS Booster up to 3,808 NVIDIA A100 GPUs, attaining 140.8 petaFLOPS in mixed precision (11.9%of peak at that scale). The time-to-solution for training FourCastNet measured on JUWELS Booster on 3,072GPUs is 67.4minutes, resulting in an 80,000times faster time-to-solution relative to state-of-the-art NWP, in inference. FourCastNet produces accurate instantaneous weather predictions for a week in advance, enables enormous ensembles that better capture weather extremes, and supports higher global forecast resolutions.