Identifier

etd-04182012-145423

Degree

Doctor of Philosophy (PhD)

Department

Oceanography and Coastal Sciences

Document Type

Dissertation

Abstract

Numerical phase-averaged wave models are the best option to obtain the spatial and temporal distribution of the wave energy over a large domain, such as the Gulf of Mexico. Parallel implementation of unstructured SWAN and WAVEWATCH-III were engaged in this research to evaluate the performance of third generation wave models for different conditions. Met-ocean data from a network of NDBC buoys and WAVCIS stations were used to assess the predictive skills of the wave models. Deep water wave energy dissipation formulations were carefully analyzed and modified to improve the accuracy of the bulk wave parameters. Moreover, the importance of the assumptions for choosing the high frequency cut-off and the slope of the power law for the frequency tail were highlighted by several simulations using SWAN and WAVEWATCH-III. The results show that previous underestimation of wave period reported from the WAM-3 formulation of SWAN was partially attributed to the different assumptions used on the high frequency end of the spectrum. When waves propagate to shallow water, several other processes affect the wave spectrum such as dissipation of wave energy by bed friction in non-cohesive environments. The wave model with an optimized set of coefficients for the Gulf of Mexico was used to skill assess two widely used bed friction formulations. Simulation results showed that the incorporation of sediment information in an eddy viscosity formulation led to more accurate wave hindcast than the JONSWAP formulation. The computation cost required to use the proposed formulation increased by less than 4%. The turbid plume exiting the Atchafalaya Bay system significantly influences the wave spectrum of western Louisiana coast. Using extended deployments during low and high discharge periods of the Atchafalaya River, meteorological, hydrodynamic and bottom boundary layer parameters were monitored from Tiger and Trinity Shoals. These datasets were used to evaluate the mud-wave interaction in SWAN. The numerical algorithm to solve the complex dispersion equation of SWAN was optimized. Moreover, the model was extended to incorporate the damping term in non-stationary simulations. The results show that without including the mud-effects, the high frequency waves were overestimated close to Tiger Shoal during northerly winds.

Date

2012

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Roberts, Harry

DOI

10.31390/gradschool_dissertations.1918

Download

Share

COinS