Hydrological evaluation of open-access precipitation and air temperature datasets using SWAT in a poorly gauged basin in Ethiopia

• Published in: Journal of hydrology (Amsterdam) 2019-02, Vol.569, p.612-626
• Main Authors: Duan, Zheng, Tuo, Ye, Liu, Junzhi, Gao, Hongkai, Song, Xianfeng, Zhang, Zengxin, Yang, Lei, Mekonnen, Dagnenet Fenta
• Format: Article
• Language: English
• Subjects: Blue Nile; Climate Forecast System Reanalysis; Climate Hazards Group InfraRed Precipitation with Station data; Satellite precipitation; SWAT; Tropical Rainfall Measuring Mission
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2018.12.026

•Open-access climate datasets were evaluated by hydrological simulation in Ethiopia.•CFSR air temperature data yielded almost identical performance to measured data.•Among three open-access precipitation data, CHIRPS had the best performance. Precipitation and air temperature are key drivers of watershed models. Currently there are many open-access gridded precipitation and air temperature datasets at different spatial and temporal resolutions over global or quasi-global scale. Motivated by the scarcity and substantial temporal and spatial gaps in ground measurements in Africa, this study evaluated the performance of three open-access precipitation datasets (i.e. CHIRPS (Climate Hazards Group InfraRed Precipitation with Station data), TRMM (Tropical Rainfall Measuring Mission) and CFSR (Climate Forecast System Reanalysis)) and one air temperature dataset (CFSR) in driving Soil and Water Assessment Tool (SWAT) model in simulation of daily and monthly streamflow in the upper Gilgel Abay Basin, Ethiopia. The “best” available measurements of precipitation and air temperature from sparse gauge stations were also used to drive SWAT model and the results were compared with those using open-access datasets. After a comprehensive comparison of a total of eight model scenarios with different combinations of precipitation and air temperature inputs, we draw the following conclusions: (1) using measured precipitation from even sparse available stations consistently yielded better performance in streamflow simulation than using all three open-access precipitation datasets; (2) using CFSR air temperature yielded almost identical performance in streamflow simulation to using measured air temperature from gauge stations; (3) among the three open-access precipitation, overall CHIRPS yielded best performance. These results suggested that the CHIRPS precipitation available at high spatial resolution (0.05°) together with CFSR air temperature can be a promising alternative open-access data source for streamflow simulation in this data-scarce area in the case of limited access to desirable gauge data.