Webinar Date : October 8, 2026
Improving convective-scale skill in regional AI weather models requires better representing rare high-impact events, broadening variable output, training at native resolution and demonstrating performance in real-time settings. Here, we evaluate these challenges using HRRRCast, a data-driven convection-allowing emulator trained on the High-Resolution Rapid Refresh (HRRR) analyses data.
HRRRCast combines native 3-km training, moderate sigmoid-weighted loss reweighing for reflectivity and precipitation, and broader system improvements including expansion of the training archive from 3 to 9 years (2016–2025), increased vertical resolution, expanded outputs, and architectural upgrades. The weighted loss improves high-threshold skill while preserving common-event behavior. Native-resolution training avoids aliasing from 6-km subsampling and substantially improves fractions skill score at convective scales, with gains that persist across pooling windows.
HRRRCast has also been running in near-real-time mode for the past year. In six months of independent observation-based verification, HRRRCast ensemble means improve several near-surface metrics relative to operational HRRR and maintain strong precipitation and reflectivity performance.
Single-member reflectivity skill is strongest at 20–30 dBZ; at 40 dBZ, skill is near parity at 6-km pooling and better at pooling radii >30 km. Hourly precipitation skill is competitive through 1.00 in and improves further in the best recent configuration. A focused comparison with HREF further shows useful low-cost probabilistic guidance for 3-h precipitation.
At storm scale, HRRRCast exhibits a displacement–structure trade-off, with generally better storm placement than storm morphology.
Daniel Abdi (NOAA GSL)
Daniel Abdi is a Research Scientist at CIRES/NOAA and the lead developer of HRRRCast, an AI-based emulator of the High-Resolution Rapid Refresh (HRRR) weather forecasting system. His research background spans high-order numerical weather prediction, computational fluid dynamics, and GPU acceleration of large-scale scientific models. His interest in artificial intelligence began through an early curiosity about using AI for game play during the emergence of systems such as AlphaGo and AlphaZero, eventually bringing together his professional work in numerical weather prediction and his longstanding interest in AI.