Predicting outflow ratios on rainfall estimation through the SPDE approach with R-INLA

A case study in ASAKAWA River basin flowing in the western part of Tokyo

Abstract

Geostatistical models of rainfall estimation can be applied to estimate amount of rainfall in a particular region. Since conventional computer-based models such as metrological model and storage function model carry unknown parametrical errors, geostatistical approach may then be useful to explore more enclosed region. This paper describes an application of SPDE with R-INLA to rainfall estimation and outflow ratios of ASAKAWA River basin. Our conclusion is that since recent global climate change causes disastrous amount of rainfall in short period of time, urban planning approach with infrastructure construction become more crucial to prevent flood.