Spatial and temporal trends in high resolution gridded rainfall data over India

Abstract

A high resolution (0.25°×0.25°) daily gridded dataset was analysed to study the spatial and temporal trend in various regions of India during the period of 1901–2013. The non-parametric Mann–Kendall (MK) test has been applied to the time series (1901–2013) rainfall data in order to detect the trends for both rainfall depth and number of rainy days. Sen slope estimator test was also used to find out the magnitude of the trend. In addition, Pettitt–Mann–Whitney (PMW) test was employed so as to determine the probable change year. The results of trend analysis of rainfall depth showed maximum number of grid points having increasing trend in summer months, while the dry months showed higher grid points with significant decreasing trend. Jammu and Kashmir, in particular, was found to have increasing trend in all months for all three levels of significances, i.e., 1%, 5%, and 10%, while NE (northeastern) region continuously reported significant decreasing trends. The annual trend analysis of the 30-year moving window of rainfall depth showed higher number of grid points with significant increasing trend during 1901–1950, decreasing trend during 1931–1980, and increasing trend again during 1961–2013. The PMW test for the rainfall depth and number of rainy days for 1901–2010 showed 1961 and 1974 as the most probable change point years with about 339 and 284 grid points agreeing upon it, respectively. The maximum increasing significant trends in rainfall was observed during south-west monsoon months. Recent years also reported an increase in rainfall intensity leading to probable increase of extreme events like floods and droughts.