3.1 Hindcasting Wind Waves and Storm Surge for Hurricanes Irma and Maria for Puerto Rico and the U.S. Virgin Islands: IOOS Test Bed Model Validation and Inter- and Intracomparisons

Monday, 7 January 2019: 2:00 PM
North 130 (Phoenix Convention Center - West and North Buildings)
Joannes Westerink, Univ. of Notre Dame, Notre Dame, IN; and B. Joyce, A. Van der Westhuysen, J. Gonzalez, and D. Yang

Hindcasts of waves and storm surge for Hurricanes Irma (2017) and Maria (2017) are examined and compared to wave and water level gauge data in the vicinity of Puerto Rico and the U.S. Virgin Islands. The region is characterized by adjacent deep ocean water, narrow shelves, and protecting coral reef systems. The storm physics is analyzed using the ADCIRC+SWAN coupled circulation and non-phase resolving wave models with respect to tides, winds, atmospheric pressure, waves, wave radiation stress induced set up, and baroclinically forced processes. The water level response is dominated by the pressure deficit to which is added wind driven surge in the shallower shelf regions to the east of Puerto Rico and wave radiation stress induced set up in near shore regions with large shore perpendicular directed waves. Geostrophically induced set up is not important. Cold water upwelling after passage of Hurricane Maria caused a 20 cm plus depression of water levels adjacent to the north coast of Puerto Rico that rebounded over the following month.

A range of wind products including synthetic vortex winds, CSFv2, HWRF, and Oceanweather Inc. (OWI) data assimilated winds are analyzed and the associated hydrodynamic responses and errors quantified. The OWI winds performed significantly better than other wind products with the CFSv2 winds aliasing the vortex, the HWRF winds generating too large a vortex with a low bias on the pressure deficit, and the synthetic vortex winds performing poorly away from the eye. The ADCIRC+SWAN and SLOSH+WAVE models are inter compared and analyzed using the best OWI winds and their response functions to each forcing component are examined. Strengths and shortcomings of each of the four component models are discussed and quantified.

- Indicates paper has been withdrawn from meeting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