S249 A Detailed Dataset of Surface Winds Over Puerto Rico from a WRF Simulation of Hurricane Maria (2017)

Sunday, 6 January 2019
Hall 4 (Phoenix Convention Center - West and North Buildings)
Jimmy Yun Ge, Rosenstiel School of Marine and Atmospheric Science, Miami, FL

The Category 4 winds produced by Hurricane Maria (2017) upon landfall in Puerto Rico resulted in widespread devastation across the island, warranting a high-resolution and accurate model simulation of the surface wind fields for further research. Achieving the closest possible representation of the actual surface wind field requires extensive validation of the simulated wind fields with land observations, yet such validation studies have rarely been performed on hurricane models, which in fact are not currently used to forecast real-world hurricane wind fields over land.

The Weather Research and Forecasting (WRF) community model, Version, is used to generate a series of high-resolution simulations of Hurricane Maria over Puerto Rico with grid spacings down to 1 km. The trial-and-error validation process consists of comparing simulated track and intensity to best-track data, as well as comparing radial profiles of total, tangential, and radial wind speeds to those calculated from independently developed real-time and post-storm surface wind analyses. The simulations that best match the observed data are repeated with higher temporal resolution to produced simulated time series of winds at fixed points. These are then compared to available wind measurements taken at weather stations across Puerto Rico before the instruments failed.

Because Hurricane Maria’s structure was also influenced by interactions with the Cordillera Central mountain range, the storm presents an interesting case of hurricane-force winds over a small surface area with diverse terrain including mountains, vegetation, and urban environments. The results of this validation study may lead to improved parameterizations of land use types in WRF, as well as improved wind speed estimates over land in hurricane models.

- Indicates paper has been withdrawn from meeting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