Downscaling Climate Model Outputs for Estimating the Impact of Climate Change on Water Availability over the Baro-Akobo River Basin, Ethiopia

Abstract

This work analyses the impact of climate change on the water resources of one of the remote basins of Ethiopia (Baro-Akobo). The study area (76,000 km2) contributes water to the White Nile in Sudan and it is one of the productive agricultural areas in Ethiopia. Currently large scale private farms are under operation in the lower basin of Baro-Akobo relying on the water resources of the basin. However, a climate change impact assessment in the basin is missing especially at the catchment scale. In this study, large scale global and regional climate model output is downscaled statistically to meteorological variables at the point scale in a daily resolution to assess future variables under climate changes. The results show that the regional model REMO is able to reproduce the precipitation and temperature observed over the basin. We noticed that statistical downscaling smooths out the bias between climate models. <br /> Hydrological models were applied to simulate current behaviour and the impact of climate change on the discharge of the SORE watershed (1711 km²), a subbasin of the Baro-Akobo river basin. A physically-based distributed hydrological model and a conceptual model were successfully calibrated and validated for recent years (1990-1997). The performance of the two models WaSiM-ETH and HBV-Light to simulate discharge were evaluated by comparing model results with measured discharge of the watershed at a daily time step. The overall performance of both models was quite reasonable at a daily scale (WaSiM-ETH R2=0.79, NSE=0.78; and HBV-Light R2=0.85, NSE=0.84). Only small differences in the performance of the two models were observed. Future discharge simulations (2011-2050) were performed using statistically downscaled climate data with both models. Both results show a similar shift of the wet season (May to July) and a reduction of the peak discharge in August and September. However, HBV-Light shows discharge increases in June by 30% to 40% while WaSiM-ETH simulates an increase up to 60% for both climate models. Therefore, methods and results described in this study could be used as indicative and supportive information on the likely impact of climate change in the Baro-Akobo basin. <br /> The onset and the cessation of the rainy season over the period of 1972-2000 and the analysis of the dry spells during the rainy season at a daily scale were analysed for the Upper Baro-Akobo basin. The results show that the stability of the onset dates of the rainy season reveals that rain-fed agricultural activity in the study area is possible at low risk. In addition, the Mann-Kendall test was applied to the Upper Baro-Akobo basin to investigate recent trends of the number and lengths of dry spells. The analyses show that there is currently no significant recent trend across the basin.