Abstract:
Watershed hydrology is responsive to changing climatic patterns. Sustainable resource planning and management at watershed scale therefore requires consideration of climate change impacts, which may result in future changes in water resources quantity and quality. Climate change is a major threat to agriculture-based livelihoods and natural resource management regimes in East Africa. Communities which rely mainly on climate-sensitive sectors for their livelihoods are the severely affected by impacts of climate change.
Mt. Elgon, which is the study area for this project, is a key water tower endowed with a rich diversity of flora and fauna which influence lives and livelihoods of thousands of people through provision of ecosystem services. The main objective of this study was to assess potential impacts of climate change on the hydrology of the four main sub-watersheds on the eastern slopes of Mt. Elgon namely Kuywa, Kimilili, Rongai and Koitobos. More specifically, this study sought to (i) calibrate and validate SWAT model which could be used to simulate the hydrological processes in Mt. Elgon sub-watersheds, (ii) simulate climate change scenarios reflecting expected rainfall and temperature change for the 2011-2040 (2020s), 2041-2070 (2050s) and 2071-2100 (2080s) periods, and (iii) assess potential hydrological impacts of the climate change scenarios.
In this study, the 2009 version of the SWAT model was used. 29-year long records (1970-1998) of daily hydro-climatic data in the upper Nzoia River basin were utilized for model setup. Model calibration was conducted using the sequential uncertainty fitting version 2 (SUFI-2) algorithm in the SWAT CUP software. Statistical downscaling was used to develop climate scenarios in form of additive and ratio anomalies between the baseline period (1960-1990) and three future periods: 2020s (2011-2040), the 2050s (2041-2070) and the 2080s (2071-2100).
The SWAT model performed well in simulating monthly stream flow with R2 values of 0.68 and 0.70 and NSE values of 0.58 and 0.70 for the calibration (1986-1998) and validation (1973-1985) periods, respectively. The delta-change method was used to generate monthly temperature and precipitation change scenarios for the future periods based on output from ten GCMs and three emissions scenarios (A1B, A2 and B1). Different magnitudes of change in climate showed varied streamflow responses in the four watersheds. The results indicated that annual rainfall is likely to increase by between 1.4% and 4.6% by the 2020s, 3.3% and 6.4% by the 2050s, and 7.3% and 15.1% by the 2080s. Projected monthly changes varied strongly depending on GCM, GHGs emissions scenario and time period. Results on streamflow response suggested potential dramatic changes in streamflow. The overall relative changes in annual mean flow in the whole catchment ranged from -2.9% to +8.9% by 2020s, -9.3% to -0.2% by 2050s and 1.7% to 22.6% by 2080s depending on the emission scenario. This implies a likely increase in floods in the area in the 2020s and 2080s. At sub-basin level, the streamflow response to precipitation and temperature change was nonlinear.
In conclusion, the sub-watersheds depicted distinct variation in their response to climate change despite their proximity.The projected climate change scenarios and streamflow change depicted wide uncertainty. However, there is need to compare results from different climate and hydrological models as well as more detailed model calibration to minimize on the parameter uncertainty. To counter the potential hydrological impacts, there is need for local and regional policy to facilitate mitigation and adaption to climate change impacts. This includes policies on water resource abstraction, flood management as well as disaster preparedness and mitigation. These policies should also include community livelihood improvement and sustainability, forest management as well agricultural managementcomponents. In addition, infrastructure development in the area should considerdetailed analysis of the potential climate changes to mitigate impacts.