ADVANCES IN THE PARAMETRIC TROPICAL CYCLONE RAINFALL MODEL, P-CLIPER: EXPANDED CALIBRATION AND ASYMMETRY IMPLEMENTATION UTILIZING DATA ANALYSIS

Abstract

While storm surge is considered the most dangerous aspect of landfalling tropical cyclones, flooding due to heavy rain associated with tropical cyclones does not fall far behind. Rising sea levels, projected increases in tropical cyclone related rainfall, and widespread urbanization are all factors exacerbating the threat of compound coastal flooding events, yet coastal communities in the United States continue to grow. Developing adequate mitigation and adaptation strategies necessitates modeling not just the risk, but the events themselves. As part of a coupled-model system, P-CLIPER is a parametric tropical cyclone rainfall model that prioritizes speed and efficiency for ensemble applications. In its current form, P-CLIPER generates axisymmetric rainfall in a 350 km radius out from the tropical cyclone storm center, but this oversimplification does not accurately capture the observed rainfall associated with tropical cyclones. P-CLIPER also utilizes a free parameter in its equations known as f which represents the change from the average rainfall intensity to allow for multiple simulations at varying intensities, but this f-value must be calibrated using historical tropical cyclones. The first objective of this thesis focuses on the calibration of P-CLIPER’s f-value, while the second objective is the development and testing of different methods for implementing an asymmetrical component to P-CLIPER. An analysis of the calibrated f-values and their behavior shows that certain tropical cyclone characteristics could be used to narrow the range of f-values for ensemble applications. This comes alongside the successful development and evaluation of an asymmetric version of P-CLIPER that shows substantial improvement over the current, axisymmetric P-CLIPER.