Study on the Continuous Simulation of Storm Rainfall Process

• Main Authors: Chin-Lung Wu, 吳進龍
• Other Authors: Ke-Sheng Cheng
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/27417669711396251552

碩士 === 國立臺灣大學 === 生物環境系統工程學系暨研究所 === 90 === This study simulates continuous storm rainfall process by using stochastic simulation. We selected hourly rainfall data of 12 raingauges in Cho-Shuei River Basin. All data are divided into 4 different rainfall types: frontal rain, Mei-Yu, convective storm and typhoon. These data are individually checked by their statistical properties and then we can simulate the whole rainfall process. First we take every 4 hours as the inter-event time to separate the whole events. Next we test the probability distribution of inter-arrival time, rainfall duration, and total rainfall depth in four different rainfall types. The inter-arrival time is found to be exponentially distributed, so we use Poisson process to simulate the points along the time frame axis during the appointed period of time. Furthermore, both of rainfall duration and total rainfall depth are also found to be exponentially distributed, and there is a positive correlation between them. Hence, we use joint bivariate exponential distribution to simulate the rainfall duration and total rainfall depth of each event. Finally, taking the non-stationary rainfall process, we use first- order Gaussian Markov process to distribute the total rainfall depth into each time interval in rainfall duration.