Date of Award

12-2025

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

School

Ocean Science and Engineering

Committee Chair

Dr.Diana Bernstein

Committee Chair School

Ocean Science and Engineering

Committee Member 2

Dr. Jerry Wiggert

Committee Member 2 School

Ocean Science and Engineering

Committee Member 3

Dr. Denis Wiesenburg

Committee Member 3 School

Ocean Science and Engineering

Committee Member 4

Dr. Kemal Cambazoglu

Committee Member 4 School

Ocean Science and Engineering

Committee Member 5

Dr. Brian Dzwonkowski

Abstract

High waves in the Gulf of Mexico, caused by tropical cyclones and winter storms, pose significant threats to coastal communities and marine activities. Wave prediction depends on precise wind field estimates, but current atmospheric model outputs often show considerable bias during severe weather events. This dissertation aims to develop and apply wind-wave modeling frameworks to obtain more accurate results and improve coastal hazard predictions in the Gulf of Mexico.

The research uses three interconnected studies with numerical modeling approaches. First, available wind data were evaluated, and wave sensitivity analysis was performed using the unstructured SWAN model with five distinct wind data sources from ECMWF and NCEP global and regional models, validated against National Data Buoy Center observations. Second, the high-resolution Weather Research and Forecasting (WRF) model was optimized through systematic testing of different domain configurations and various physics parameterization schemes during Hurricane Michael (2018). Third, a multi-physics ensemble simulation framework was adopted by integrating the WRF model results with SWAN wave simulations to simulate hurricane-induced waves. Key results show that FNL and NAM wind datasets offer better correlation with observations, while ECMWF Real-Time products combined with Janssen schemes give the best wave estimates, with minimal error and maximal correlation. The selected nested domain setup at 0.1° resolution cut the RMSE by up to 10% and reduced bias by over 60% compared to other available datasets. The best WRF physics setup includes Lin and WSM6 microphysics with the YSU planetary boundary layer scheme, and performed consistently across coastal stations, with fair agreement at the nearshore stations with higher complexity and more importance. The ensemble simulation framework, using eleven WRF simulations with varied wind inputs for SWAN, proved more reliable through simple averaging and an ensemble wind-related product. Overall, these findings help recommend a framework for high-precision coastal wind-wave prediction during extreme tropical cyclone events. By optimizing wind field generation and wave modeling, this research may offer an essential tool for wind-wave prediction, reliable for coastal management, marine planning, and engineering resilience. The ensemble simulation approach has great potential for operational use for hazard warning systems, supporting better risk reduction and public safety in the Gulf of Mexico coastal areas, especially for the Mississippi Bight.

ORCID ID

0000-0002-3448-2103

Available for download on Tuesday, August 01, 2028

Share

COinS