Author(s): Thottungal Drissia, Vinayakam Jothiprakash, Alayil Bahuleyan Anitha
Linked Author(s): T K Drissia
Keywords: L moments, generalised logistic, generalised pareto, geographic information system
Abstract: Flood frequency analysis is a prerequisite for designing flood control structures. It is also used in flood management, design floods and flood investigation works. Gumbel method is commonly used for predicting peak discharge for a given return period. Nevertheless, this method gives at site flood frequency for a given return period. In this study, probability weighted moments and L moments were used for estimating the at site flood frequency. Probability weighted moments was used for estimating the parameters for Gumbel Extreme Value distribution and Log Pearson Type III distribution at a given site. The study area taken is the west flowing rivers in Kerala, India. The whole area is 40849 km2 with 41 small rivers originating from Western Ghats and reaches Arabian Sea. L moments were used for the parameter estimation of Generalised Extreme Value (GEV), Generalised Pareto (GPA) and Generalised Logistic (GLO) distribution. The best fit distribution was found using statistical indices. Using the best fit distribution, flood frequency was calculated for all the gauging stations like isohyets and isopluvial lines, an attempt was made to create contour of equal peak discharge for return periods 20, 50 and 100 years. This was achieved using ArcGIS 10, GIS software. These curves give information about the magnitude of flood frequency in each location
Year: 2017