Identification of Rainfall Trends over Four Cities of Assam, India

Abstract

In time-series analysis, several non-parametric and parametric techniques are available to identify a trend. However, due to the lesser sensitivity of the non-parametric methods towards the outliers, they are often given priority over parametric methods for analysing the trend. In this study, the significance of the trend in the rainfall data of the four cities: Guwahati, Tezpur, North Lakhimpur, and Dibrugarh in Assam is examined using the non-parametric Mann–Kendall method. The non-parametric Sen’s estimator method is used to estimate the magnitude of the significant trends for the four cities. The rainfall trend has been studied based on monthly, yearly, and seasonal observations. The results clearly show that there is no trend in the cities except for Dibrugarh, where an increasing rainfall trend is observed within 5% significance level. Based on the seasonal rainfall analysis, it is noted that the pre- and post-monsoon precipitation has a significant increasing trend, while no trend is seen during the monsoon period, which indicates an absence of significant change in the rainfall magnitude during the monsoon for the study period. The analysis of the monthly rainfall data has reveal a significant trend for July in the cities of Guwahati and Tezpur, which are closely located, while the significant trend for May is seen in the city of North Lakhimpur. However, for Dibrugarh, all months have shown significant trends at 5% except for September, November, and December. The analysis has shown that there is a significant change in the rainfall trend for the non-monsoon period, indicating the change in the rainfall pattern, while the overall trend for the monsoon period remains the same. This change of rainfall pattern during the non-monsoon period may be an impact of the climate change, which requires detailed studies. The overall analysis for the four sites over the period under review has shown a significant trend in rainfall only for Dibrugarh.