Compound floods, which can be attributed to different drivers (pluvial, fluvial, surge, tide, and waves), generate a larger flood hazard when drivers co-occur than when they occur in isolation of each other. Current compound flood risk assessments are affected by a curse of dimensionality, where a larger number of events need to be numerically simulated to understand the response of risk to drivers. This research aims to create a methodology that improves the quantification of compound flood risk by using a Treed Gaussian Process (TGP) for the case study of Charleston. A TGP can actively learn from the response of damages to drivers to reduce the number of events that need to be simulated. By comparing this approach with a state-of-the-art approach, the research shows a reduction of the computational cost by a factor of 4, an improvement in the root mean square error by a factor of 8, and an improvement in the estimate of Expected Annual Damages (EAD) by a factor of 20. This reduction in computational cost allows for the inclusion of random variables that are normally assumed constant such as the duration and time lag of drivers. A sensitivity analysis demonstrates these variables produce a statistically significant difference in the estimate of EAD, which increases its value from 172 to 219 Million USD. The research also shows the combination of events caused by drivers leading to extreme damage changes when including these additional random variables, although surge is always found to be dominant. By applying the TGP to multiple outputs, the research demonstrates the TGP is not only applicable to the case study, which shows a TGP can be implemented in current flood risk assessments.

Van Sickle Island, located in the Suisun Marsh in the San Francisco Bay Area, has experienced multiple levee breaches in the past. Due to the large social and economic implications of flooding, questions have arisen about whether the current management of the island is still feasible. This project, therefore, aimed to find the most financially favorable management plan for the upcoming 50 years. First, an understanding of the system was obtained through a site investigation, a multivariate analysis, and a hydrodynamic model. It was concluded that discharge, tide, air pressure, and wind setup all contribute to the water level. These variables were then used as input for a 1-dimensional hydrodynamical model, which quantifies their effect on the water level. Subsequently, four management plans were considered: status quo, raising the levees, conversion into an estuarine wetland, and abandoning the island. To allow for comparison between these management plans, a cost-benefit analysis and a net present value calculation were performed for every plan. One of the major contributors to the costs is risk. Risk is defined as the product of the probability of failure and the consequence of that failure. The failure mechanism assessed is overflow, as this is the most relevant one. To quantify this failure probability, a statistical model was developed. This model includes an extreme value analysis and an event tree. Due to large uncertainties in the behavior of both the levee and the water level over time, the original 50-year lifetime of the management plan was deemed too ambitious. Therefore the lifetime was reduced to 10 years. The conclusion of the cost-benefit analysis and the net present value was to convert Van Sickle Island into an estuarine wetland. This is because it was the only management plan that was profitable after 10 years. The net present value was found to be equal to $14,924,048.