Spatial homogeneity of extreme precipitation indices using fuzzy clustering over northeast India

Abstract

Regionalization on the basis of the properties of hydro-meteorological data helps in identifying the regions reflecting the similar characteristics which could be useful in designing hydrological structures as well as planning and management of water resources of the region. In this study, rainfall data of northeast India were utilized for calculation of extreme precipitation indices as suggested by expert team on climate change detection and monitoring. Trend analysis of the indices was carried out using Mann–Kendall nonparametric test, and Sen’s slope estimator was used for calculating the magnitude of trend. Further fuzzy c-means method was used for clustering of the selected stations on the basis of six parameters of all these precipitation indices, i.e., latitude, longitude, mean, standard deviation, minimum value and maximum value. Three cluster validity indices, namely fuzzy performance index, modified partition entropy and cluster separation index were used for selecting the optimum cluster numbers. Analysis shows insignificant trend for the indices like consecutive dry days and consecutive wet days, whereas maximum 1-day precipitation (R1 day) and maximum 5-day precipitation (R5 day) are not showing any clear trend. It is observed that the number of rainy days is decreasing followed by increasing 1-day precipitation. Cluster analysis of the precipitation indices shows five major clusters for most of the indices. © 2019, Springer Nature B.V.