Rainfall prediction for landfalling tropical cyclones: perspectives of mitigation

The torrential rainfall associated with landfalling tropical cyclones (TCs) often represents the major impact to coastal regions, but at the same time an enormous challenge to meteorologists and forecasts. This chapter first discusses the complex dynamical processes involved in TC landfalls, which are related to the increased surface roughness and reduced surface moisture fluxes of land. The result is often certain patterns of convection and rainfall asymmetry in the landfalling TCs, but these patterns are not well explained by current theories or conceptual models. With emphasis of development of rainfall prediction techniques according to the needs of mitigation, the requirements on the skill of rainfall forecasts from the perspectives of mitigation are reviewed. Then, the operation and performance of several statistical TC rainfall models are discussed including the rainfall climatology-persistence model (R-CLIPER) for the Taiwan area. A topographic component is developed for R-CLIPER through multiple regression analyses, which improves the model's performance in reproducing the local extreme rain that is lacking in the original model. Finally, the importance of utilizing remote-sensing data in TC rainfall prediction is discussed, and how TC rainfall statistical models can be applied to risk analyses under the consideration of global changes.