Poster Session P1.63 HWRF performance diagnostics from the 2009 Atlantic hurricane season

Tuesday, 11 May 2010
Arizona Ballroom 7 (JW MArriott Starr Pass Resort)
Brian D. McNoldy, CIRA/Colorado State Univ., Fort Collins, CO; and M. DeMaria, V. Tallapragada, and T. Marchok

Handout (1.0 MB)

The performance of the Hurricane Weather Research and Forecast (HWRF) model during the 2009 Atlantic hurricane season was less than optimal, particularly for storms in a vertically-sheared environment. A common vertical shear metric is the vector difference between the 850 hPa winds and the 200 hPa winds averaged around the storm to remove the storm's circulation contribution. This metric, along with several others, is calculated from operational HWRF output and from Global Forecast System (GFS) analyses as validation. The other parameters included in the diagnostic output include storm center location, maximum 10-m wind speed, sea surface temperature, and distance to land. During the 2009 season, HWRF often had a high intensity bias, meaning that its forecast intensities were consistently too high. Even in cases where it represented the vertical shear reasonably well, the model vortex did not respond to that shear realistically. Alternatively, there were cases where the vertical shear was under-represented in the model, and the model vortex intensified too much owing to the perceived weak shear. The Geophysical Fluid Dynamics Laboratory (GFDL) model performed more realistically and will also be utilized to help identify biases in the HWRF model.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner