Given the past history and future risk of storm surge in the United States, alternative storm protection techniques are needed to protect vital sectors of the economy and population, particularly within southeastern Louisiana. It is widely hypothesized that coastal wetlands offer protection from storm surge and wave action, though the extent of this protection is unknown due to the complex physics behind vegetated flow dynamics. This thesis presents numerical modeling results that estimate the relative sensitivity of waves and storm surge to characteristics embodied by coastal wetlands. An idealized grid domain and 400 km2 (20 km by 20 km) marsh feature provide a controlled environment for evaluating marsh characteristics, including bottom friction, elevation, and continuity. Marsh continuity is defined as the ratio of healthy marsh area to open water area within the total wetland area. It is determined that increased bottom friction reduces storm surge levels and wave heights. Through the roughening of the bottom from sandy to covered with tall grass, it is estimated that waves may be dampened by up to 1.2 m at the coast, and peak surge may be reduced by as much as 35%. The lowering of marsh elevation generally increases wave heights and decreases surge levels, as expected. A 3.5 m decrease in marsh elevation results in as much as a 2.6 m increase in wave height, and up to a 15% decrease in surge levels. Reductions in marsh continuity enhance surge conveyance into and out of the marsh. For storms of low surge potential, surge is increased by as much as 70% at the coast due to decreasing marsh continuity from 100% to 50%, while for storms of high surge potential, surge is decreased by 5%. This indicates that for storms of high surge potential, a segmented marsh may offer comparable surge protection to that of a continuous marsh. Wave heights are generally increased within the marsh due to the transmission of wave energy through marsh channels. Results presented in this thesis may assist in the justification of coastal wetland mitigation, and optimize marsh restoration in terms of providing maximum storm protection.

Wave models are an integral part of coastal engineering due to their ability to quantify information that is either unobtainable or unavailable. However, these models rely heavily on the input of a directional wave spectrum that describes the variation of energy with frequency and direction. This study investigated how five methods for computing the directional wave spectrum perform within the nearshore spectral wave model, STWAVE. The results of the five experimental runs showed that overall, the greatest differences between spectra were observed in the significant wave height parameter. The mean wave direction showed greater differences at the offshore model domain boundary and lesser differences as the wave enters the nearshore; and the peak period had fewer differences at the boundary, but at the nearshore the differences were dependent upon the presence of wind forcing. Winds had a significant impact on observed differences between the spectra in the domain by dominating the wave field variation.

The need for a consistent and accurate production of long-term nearshore wave record is discussed. With multiple decades of offshore hindcasts and long, continuous data sets available, it is possible to create a nearshore synthetic wave record. The Brevard County coastline offers an area with a high quality 62-year long offshore hindcast, as well as an 11-year long and nearly continuous high-resolution nearshore wave record to compare with model performance. This thesis presents the steps in the development and performance of the synthetic nearshore wave record produced. A novel approach was used to compare, validate and calibrate this type of data which included using quantile-quantile plots and bin-averaged scatter plots. In a comparison between two reputable deep-water hindcasts (MSC50 and Wavewatch III), it was found that Wavewatch III significantly underpredicts wave heights in the higher range (>8m). At the nearshore STWAVE proves to be a simple, robust and fast way to create a nearshore wave record. Root mean squared error (0.272m-0.317m) and modified index of agreement (0.697-0.646) values for significant wave height show promising results for overall model performance with the currently available hindcast. Possible future improvements could be made by modifying the offshore hindcast to have finer grid resolution and further studying different friction models for the nearshore wave transformation model. Overall, the use of the MSC50 hindcast, to drive STWAVE at the nearshore, exhibits good agreement with ADCP data and analysis for significant wave heights can be used with confidence. Currently, no long-term trends can be resolved with the available record at the location used herein, yet more years of data/hindcasts in the future could provide more evident trends in wave climate change.