Rainfall trend and variability analysis of the past 119 (1901-2019) years using statistical techniques: A case study of Kolkata, India

Abstract

The core purpose of this study is to investigate the spatial variation in monthly, seasonally, and yearly rainfall patterns in the Kolkata district of West Bengal, India, between 1901 and 2019. (Around 119 years). The trend's reliability and intensity were assessed non-parametrically by applying monthly rainfall data series and the Mann–Kendall and Sen's slope estimators. The data showed a considerable increase in pre-monsoon, monsoon, post-monsoon, and also annual rainfall while decreasing in winter rainfall across the district of Kolkata. The positive trend is identified in the data series of pre-monsoon, monsoon, post-monsoon, and annual rainfall, however, winter rainfall exhibited negative trends. The highest increase in rainfall was observed during the post-monsoon season (0.365091 mm year-1), with the smallest increase (0.232591 mm year-1) occurring during the pre-monsoon season. In the winter season, there was a notable rain that has declined substantially(-0.01356 mm year-1). The coefficient CV, %, was used to determine the pattern of rainfall variability. The winter rainfall exhibited the highest CV rating (72.89%), but annual rainfall showed a minimum CV value (17.68%). Generally speaking, a high variance in CV was discovered, indicating that the whole area is very sensitive to droughts and floods. For future forecasts, there is a considerable difference in monthly rainfall data between linear regression and SMOreg, while the annual rainfall is little difference between linear regression, SMOreg, and CA-ANN analysis.