|
Abstract:
|
Properly assessing the storm surge risk is important for coastal areas. Typically, we obtain the simulation about water level response towards a collection of synthetic hurricane storms (input distribution) via statistical models. In this scenario, it is crucial to have a good estimate of the input distribution for the simulation. We attempt to get better storm surge input distribution for measuring the model of storm evolution. However, the current practice is ad-hoc and does not account for the uncertainty of the estimation. In this research project, we investigate various methods to improve the estimation efficiency and to properly quantify the uncertainty. We demonstrated the exploratory analysis of synthetic storm tracks, and compared the simulated tracks with the real storm data. One of the main discussions in this project is how to improve the current practice with the technique of spatial statistics, such as using hierarchical model to improve the estimation of input distribution, or spatial-temporal point process modeling for the storm occurrence rate. We also fusion different source of data to offset the limitation of insufficient real hurricane observations.
|
