Return Period Analysis of Major Flood Events Considering Homogeneous Regions

Abstract

Extreme flood events cause huge amount of money and human resources in handling the catastrophic during and post to the disaster. In order to cope for these events, mitigation strategies to reduce the impact or occurrence are crucial. One of the ways is to better design the flood defense structures using possible highest rainfall values that might occur at a specific location represented by rainfall return periods. The main idea of this research is to consider using extreme rainfall homogeneous regions to help predict the best return period for designing flood defense structure using information of historical flood events occurring at the location of similar extreme rainfall homogeneous regions. The objectives of this research are to identify major flood events in Peninsular Malaysia and to map the rainfall return period associated to the flood events within similar homogeneous regions. The method of pooled frequency analysis was used to combine annual maximum series (AMS) data in each homogeneous region. The best distribution model is selected based on highest rank according to Kolmogrov Smirnov, Anderson Darling, and chi-squared. The highest return period of rainfall event occured at zone 1, zone 2, zone 3, and zone 4 is 50 years, 333 years, 500 years, and 20 years, respectively. The use of the highest return period values may be applied for designing flood defense structures located in the specific homogeneous region as those return periods have already occured within respective homogeneous regions.