Rainfall Variability and Trends in the Lower Dhansiri River Catchment, Assam, India

Abstract

The Lower Dhansiri River catchment in Assam, India, is one of the highly sensitive hydroclimatic zones in northeastern India, where rainfall patterns critically influence river discharge, agriculture and socio-economic systems. This study investigates long-term rainfall variability and trends from 2001 to 2024 using daily rainfall data from meteorological stations in the catchment. The analysis incorporates annual, seasonal (monsoon and post-monsoon), event-based and storm-based rainfall characteristics, applying statistical parameters such as mean, standard deviation, coefficient of variation (CV), skewness, kurtosis and rainfall intensity. Results reveal substantial inter-annual variability with a mean annual rainfall of 1,222.4 mm and a CV of 120.9%, indicating high year-to-year fluctuations. The monsoon season (May–September) contributes approximately 84% of annual rainfall, with extreme events such as in 2010 (2884 mm) and 2009 (650.1 mm) highlighting increasing extremes. Post-monsoon rainfall, particularly in October, shows erratic behavior, with a notable extreme in 2010 (1862.7 mm), reflecting possible shifts in weather systems. Event-based analysis indicates that most rainfall occurs in short-duration events, with one-day events contributing 14.29 mm per event. Z-score classification of storm events reveals a dominance of extremely low (915 rainy days) and very low intensity events, while very high intensity events are rare but catastrophic. Skewness and kurtosis values consistently show positive skew and leptokurtic distribution, confirming the presence of extreme rainfall events. The study underscores the intensification of rainfall extremes and the increasing frequency of anomalous events, likely influenced by climate change, land-use change and regional atmospheric dynamics. These findings have critical implications for flood risk management, agricultural planning and water resource sustainability in the region.