Impacts of Rainfall Variability on Streamflow in the Drylands of Northern Kenya: Assessing Water Availability under a Changing Climate

Abstract

Water availability is a function of climatic and land surface conditions, which determine the amount and distribution of atmospheric water as it reaches the surface. This largely depends on rainfall, whose variability affects water, food and livelihood security. This paper sought to quantify the effects of rainfall variability on water availability in an effort to support effective water resources management. Coefficient of Variation (CV), Standardized Anomaly Index (SAI) and Mann-Kendall trend test approaches were used to assess variability and trends, while correlation and regression analysis were employed to determine effects of rainfall variability on streamflow. A hydrological model, Soil and Water Assessment Tool (SWAT), was used to simulate streamflow with a view to assessing water availability under two climate change scenarios; Representative Concentration Pathways (RCP) 4.5 and 8.5. Results show that the area experiences moderate to extreme rainfall variability, as indicated by CV and SAI values that ranged from 20 to 99% and -2.5 to +3 respectively, resulting in moderate to extreme floods and droughts that often disrupt livelihoods. Current streamflow simulations (1981 – 2020) indicated increasing trends. Near and far future streamflow volumes will decrease by 15% during the April-May-June season while during October-November-December season they will increase by 13%, compared to the present. Increasing trends of rainfall and streamflow indicate that the area has opportunities for rainwater harvesting while the high variability indicates a need for early warning systems to cushion communities from climatic shocks evidenced by the impacts of climate extremes experienced in the area.