Rainfall Analysis over the Niger Central Hydrological Area, Nigeria: Variability, Trend, and Change point detection

Abstract Rainfall data are a vital meteorological input to agricultural modelling systems and water resources planning and management studies. In this study, CRU data (CRU_TS 4.01) was used to investigate spatiotemporal variability of rainfall at 33 sub-basins of the Niger Central Hydrological Area (NCHA), Nigeria, over 105 years (1911–2015). For Rainfall variability studies, rainfall variability index, precipitation concentration index, and linear regression model (LRM) were employed. Trend analysis and change point detection were done using nonparametric Mann-Kendall (MK) tests, Standard Normal Homogeneity Test (NSHT), and Pettitt's test (PT), and for spatial analysis of rainfall, kriging interpolation method was used. Rainfall variability index showed that 1983 and 1911 were the wettest and driest years, respectively, while 1921-1930 and 1981-1990 were the wettest and driest decades, respectively. Precipitation concentration index indicated that the NCHA had experienced 61% moderate rainfall and 39% high concentration rainfall. LRM showed that the change in the mean annual rainfall, early rainy season, and the late rainy season had occurred at the rate of -0.30 mm/year, -0.27 mm/year, and -0.23 mm/year, respectively. MK results showed that except for sub-basins 21 and 22 (which are positive), all other sub-basins have a negative trend signifying a decrease in trends of rainfalls. NSHT and PT suggested 1969 as the probable change point for the whole basin. The decrease in the amount of rainfall (after the change point) indicates a drier condition over the basin. Therefore, farmers and water resources personnel need to focus attention on the management of this vital resource by making the right choice of what and when to plant and managing the water crisis (drought and floods) which emanates from erratic rainfall.