Estimation of Design Rainstorm for Rural and Urban Area Using Gumbel’s and Log-Pearson Type III Method

Abstract

Growing population and migration in Selangor and Kuala Lumpur led to changes in the usage and management of lands which affect the hydrologic cycle within the area. Every year, flood has been reported in those areas and occurred in a random and unpredictable manner. One of the alternatives to prevent future occurrence of flood is by predicting the design rainstorm, and it is the responsibility of the department of irrigation and drainage to estimate the future hydrological events. Intensity–duration–frequency curve is one of the methods to estimate the rainstorm event of a particular location. Rainfall data are needed to be obtained via installation of rain gauge at a rainfall station before the data are analyzed. In this study, four separate rainfall data of urban and rural areas at intervals of 15, 30, 45, 60, 180, 360, 720, 1440, and 4320 min were recorded to be analyzed. The data were analyzed using two commonly used methods, namely the Extreme Value Type I (Gumbel’s) and Log-Pearson Type III method. Intensity–duration–frequency (IDF) curves for every station were plotted using results obtained from both methods, and an empirical formula for the design of rainstorm was developed. Accuracy of both methods was compared by using goodness-of-fit test, which aims to identify the most suitable method to be carried out. Applications of empirical formula will be useful for engineers to design rainstorm in water resource planning in order to prevent flood in future.