9B.2 Toward Global 6.5-km Weather Prediction and Storm-Resolving Simulation

Wednesday, 15 January 2025: 8:45 AM
207 (New Orleans Ernest N. Morial Convention Center)
Linjiong Zhou, GFDL, Princeton, NJ; and L. M. Harris, J. H. Chen, K. Gao, K. Y. Cheng, M. Tong, A. Kaltenbaugh, M. Morin, and L. Johnston

A common strategy to improve weather model forecasts is to increase the model's resolution, which requires more computing power and the inclusion of updated dynamic and physical processes to handle smaller-scale features. The Geophysical Fluid Dynamics Laboratory has developed the 13-km System for High-resolution prediction on Earth-to-Local Domains (SHiELD) for global weather prediction and the 3.25-km eXperimental-SHiELD (X-SHiELD) for global storm-resolving simulation. However, the 13-km model is not sufficient for some high-impact weather systems, but the computational expense of the 3.25-km model has not been affordable for regular use. In this study, we introduce a 6.5-km version of SHiELD, designed to bridge the gap between medium-range global weather prediction and global storm-resolving simulation while remaining feasible for real-time forecasts. This model operates in the "gray zone" (at grid spacings of 10 km or below), where thunderstorms are partially resolved, necessitating adjustments to physical parameterizations originally designed for coarser resolutions. Comparative analyses with the 13-km SHiELD over a three-year hindcast period show significant improvements in global, regional, tropical cyclone, and continental convection predictions. These findings demonstrate that the 6.5-km SHiELD can be used to advance weather prediction by effectively addressing both synoptic weather systems and specific storm-scale phenomena in a single global model.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner