Towards operational data-driven regional weather forecasting: extreme-event skill and near-real-time verification with HRRRCast
Abstract
Improving convective-scale skill in data-driven regional weather models requires better representing rare high-impact events, training at native resolution, broadening variable output, 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) analysis data. HRRRCast combines native 3-km training, moderate sigmoid-weighted loss reweighting 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 seven 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. Precipitation skill remains competitive through 1 in/h and extends to about 3 in/h 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.
Related
- Operational chemical weather forecasting with the ECCC online Regional Air Quality Deterministic Prediction System version 023 (RAQDPS023) – Part 1: system description
- HourGlass: A probabilistic data-driven temporal downscaler for global and regional weather forecasting
- SDECast: Probabilistic Weather Forecasting in Continuous Time with Neural SDEs
- Swift: an autoregressive consistency model for efficient weather forecasting
- Butterfly Effect Confirmed in Global AI Weather Models: Evidence from Tropical Cyclone Forecasting
BibTeX
@misc{abdi2026towards,
title = {{Towards operational data-driven regional weather forecasting: extreme-event skill and near-real-time verification with HRRRCast}},
author = {Daniel Abdi and Isidora Jankov and Vanderlei Vargas and Paul Madden and Emily Carpenter and Travis Wilson and Jeff Hamilton and Molly Smith and Lindsay Blank and Linlin Cui and Jun Wang and Matthew Mahalik},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.36186},
eprint = {2610.36186},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.36186},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Towards operational data-driven regional weather forecasting: extreme-event skill and near-real-time verification with HRRRCast
AU - Daniel Abdi
AU - Isidora Jankov
AU - Vanderlei Vargas
AU - Paul Madden
AU - Emily Carpenter
AU - Travis Wilson
AU - Jeff Hamilton
AU - Molly Smith
AU - Lindsay Blank
AU - Linlin Cui
AU - Jun Wang
AU - Matthew Mahalik
PY - 2026
DA - 2026/10/26
PB - arXiv
DO - 10.48550/arxiv.2610.36186
UR - https://arxiv.org/abs/2610.36186
AB - Improving convective-scale skill in data-driven regional weather models requires better representing rare high-impact events, training at native resolution, broadening variable output, 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) analysis data. HRRRCast combines native 3-km training, moderate sigmoid-weighted loss reweighting 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 seven 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. Precipitation skill remains competitive through 1 in/h and extends to about 3 in/h 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.
ER -