A Method for Evaluating the Impact of Typhoon-Induced Rainfall in Taiwan under Climate Change Scenarios Based on Statistic Models
Tsun-Hua Yang(a*), Hao-Lun Kuo(a), Chi-Ying Lin(a)

a)National Yang Ming Chiao Tung University
*tshyang[at]nycu.edu.tw

Abstract

Typhoon-induced rainfall is a significant water resource for the Northwest Pacific region. The amount of rainfall from a typhoon is closely related to the duration of the typhoon^s presence and its movement speed. Future climate change scenarios suggest that rising ocean temperatures will affect the formation of typhoons. In this study, statistical models such as the Negative Binomial distribution, Vickery Model, and Monte Carlo simulation were used to analyze changes in the total number of typhoons, frequency, typhoon genesis locations, path distributions, and wind field intensity under future climate change scenarios. A Bayesian network was then developed to predict typhoon-induced rainfall. The network includes a directed acyclic graph to visually represent the relationships between typhoon-induced rainfall and typhoon characteristics such as typhoon radius, intensity, wind speed, and location. For validation of the statistical models, this study selected 30 rainfall gauges across Taiwan island and used typhoon data from 1977 to 2022 and rainfall data from 2010 to 2021. Finally, the study assessed the potential impacts of typhoon-induced rainfall on a categorized basis towards the middle and end of the century under a high-emission climate change scenario (RCP 8.5).

Keywords: Climate change- Typhoon rainfall- Statistical approach- Bayesian network

Topic: Interdisciplinary Earth Science and Technology