Statistical analysis of long-term rainfall trends in Cherrapunji, Meghalaya, India

Abstract

Rainfall is the key climatic variable that governs the regional hydrologic cycle and availability of water resources. Rainfall trend analysis in a localized watershed can improve many aspects of water resource management not only to the catchment itself but also to some of the related other catchments. The trend analysis of monthly rainfall data over Cherrapunji of Meghalaya in India for the period 1872-2007 has been carried out in this work. While the magnitude of the trend in the time series has been determined using Sen's estimator, the significance of the trend in monthly rainfall series has been tested using Mann-Kendall test. During the time span 1872-2007, an increasing trend has been found in the monthly rainfall for the months July, October and November, and a decreasing trend has been found in the monthly rainfall for the months February to June, August and September. On the other hand, it was found that none of Mann-Kendall Z values was significant at 5% level of significance. Therefore, from Mann-Kendall Z test, it can be concluded that there is no trend in any month in monthly rainfall for the station Cherrapunji. For the better assessment of the temporal variation in monthly rainfall trend, whole period was divided into two halves, 1872-1939 and 1940-2007. Then, trend magnitude through Sen's estimator and Mann-Kendall Z for test of significance were determined for these two time periods separately. The analysis of trends of monthly rainfall in these two halves showed large variability in the magnitude and direction of the trend in various months from one half to another. Accurate prediction of trends in monthly rainfall is an important aspect of climate research and we believe that present study could provide a scope to correlate between current rainfall trend and climate change scenario of the study area.