9A.1 Verification of short-term thunderstorm rotation forecasts in the 1-km Warn-on-Forecast System

Wednesday, 15 January 2025: 8:30 AM
206 (New Orleans Ernest N. Morial Convention Center)
Christopher A. Kerr, CIWRO/NSSL, Norman, OK; and P. S. Skinner, D. R. Stratman, H. E. Brooks, B. C. Matilla, Y. Wang, and N. Yussouf

Since 2021, a Warn-on-Forecast System with a 1-km horizontal grid spacing (WoFS-1km) has been under development in hopes to produce better short-term ensemble forecasts (up to 3 hours) of individual convective storms and associated hazards than the 3-km Warn-on-Forecast System (WoFS-3km). In addition to the decreased model horizontal grid spacing, WoFS-1km benefits from the assimilation of higher resolution radar and satellite observations. Recent studies demonstrate WoFS-1km has fewer spurious storms and more accurate storm motions while also detecting more small (< 400 km2) thunderstorms compared to WoFS-3km.


This study focuses on the prediction of mid- and low-level rotation (proxies for storm severity and multiple hazards) within individual thunderstorms by WoFS-1km and WoFS-3km. Mid- and low-level model updraft helicity is verified using Multi-Radar Multi-Sensor (MRMS) rotation via an object identification and matching technique within storms detected by both WoFS-1km and WoFS-3km. Rotation hits occur when both the forecast and observed storms are rotating, false alarms when only the forecast storm is rotating, misses when only the observed storm is rotating, and correct negatives when neither are rotating. From these quantities, contingency table metrics are computed. Probability of detection of rotating storms is higher in WoFS-1km for all lead times. While the false alarm ratio of rotating storms is similar in WoFS-1km and WoFS-3km, critical success index is higher in WoFS-1km. For rotation object hits, errors in forecast rotation object size, intensity, and location are also assessed.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner