Regionalization and Trend Analysis of Seasonal Precipitation (1970 - 2020) using Hierarchical Clustering and Wavelet Transform Approach in the Mahi River Basin, India

Abstract

Understanding the trend in seasonal precipitation in the context of climate change is crucial for the maintenance of regional water resources management. The present study examines the seasonal precipitation trend in the Mahi River basin, India by using the wavelet transform and clustering method. Very limited studies have been carried out over the basin using this approach. Daily gridded rainfall data (0.25°×0.25° spatial resolution) for 51 grids from 1970 to 2020 has been taken from India Meteorological Department (IMD) and the time series trend was tested by Mann-Kendall. The monthly, seasonal, and annual statistic has been analyzed for 51 years. We have also plotted the relationship of precipitation trend–elevation. After that, homogeneous precipitation regions are delineated with hierarchical clustering analysis. Results reveal that seasonal precipitation over the basin clusters into 4 subregions for monsoon (JJAS1 to JJAS4) and 3 subregions for winter, pre-monsoon, and post-monsoon seasons. After the regionalization of the subregions, the periodicity and the inter-seasonal relationship were analyzed using continuous wavelet transform (CWT). Findings revealed that the cross-correlation between pre-monsoon and monsoon seasons highly affect significantly in past over the basin. Further, the regional geography-altitude and moisture convergence perform a major role in the variability of seasonal precipitation. The output of the current study would be utilized in a better understanding of precipitation forecast, watershed, and agricultural planning which may give a better climate change indication at a regional-basin level. We also suggest future needs to more focus on the understanding of elevation-dependent wetting in hilly regions using a different type of datasets.