Authors: Xiao Wang (Oak Ridge National Laboratory); Zezhong Zhang (Auburn University); Isaac Lyngaas, Hong-Jun Yoon, Jong-Youl Choi, and Siming Liang (Oak Ridge National Laboratory); Janet Wang (Tulane University, Oak Ridge National Laboratory (ORNL)); Hristo G. Chipilski (Florida State University); Ashwin M. Aji (Advanced Micro Devices (AMD)); Feng Bao (Florida State University); Peter Jan van Leeuwen (Colorado State University); Dan Lu (Oak Ridge National Laboratory); and Guannan Zhang (Oak Ridge National Laboratory (ORNL))
Abstract: Accurate Earth system prediction requires state inference from incomplete observations, but conventional two-stage data assimilation (DA) is computationally prohibitive because repeated PDE-based ensemble forecasts, observation updates, and intermediate data movement limit ensemble size at high resolution. We introduce STORM, a one-stage generative AI framework that reformulates DA as diffusion-based posterior sampling, replacing PDE forecasts with scalable AI inference. It combines a spatiotemporal transformer with a global-attention algorithm that reduces complexity from quadratic to linear through scalable gradient propagation, enabling high-resolution, long-context Earth modeling. STORM scales to 74,400 GPUs on Frontier with 96%–99% strong-scaling efficiency and up to 6 ExaFLOPs sustained BF16 throughput, while enabling 32,768-member ensembles for uncertainty quantification in 34 seconds on 4,096 GPUs. It scales to 20 billion spatiotemporal tokens and 177,000 temporal frames. Hurricane tracking and long-term climate reanalysis demonstrate improved accuracy, including benefits from longer temporal context and recovery of temperature extremes missed by forecast-only predictions.