Streamflow Simulation by SWAT Using Different Precipitation Sources in Large Arid Basins with Scarce Raingauges

Streamflow from the mountains is the main source of water for the lower plain in arid regions. Accurate streamflow simulation is very meaningful for the prediction, planning and management of the water resources. Precipitation is one of the most important inputs for accurate runoff simulation and its detailed spatial distribution is needed, because the degradation of raingauge networks strongly affects the simulated hydrographs. The previous studies provide many methods to get the desired density of precipitation networks. However, many large arid drainage basins in northwestern China have low density of precipitation stations, which makes the streamflow modeling and prediction very difficult.

Therefore, researchers from Xinjiang Institute of Ecology and Geography taking the Manas River basin as a case study, which covers an area of 5,100 km² with only one precipitation station located, aims at improving the precision of streamflow simulation in large basins on the basis of realizing the spatial pattern of precipitation with scarce raingauges. Based on raingauge data and Tropical Rainfall Measuring Mission (TRMM) data combined with raingauge data, different approaches were explored for spatializing precipitation in this area with scarce raingauges. Spatialized precipitation was then input into Soil and Water Assessment Tool (SWAT), a semi-distributed hydrological model, to simulate streamflow.

Results of the study indicated that SWAT performed watershed simulations reasonably well using both of the proposed precipitation parameterizing methods. Visual inspection from hydrographs and statistical indicators all showed that the simulation performance of SWAT is better when using radar data combined with raingauge data. It can be seen that the accuracy of precipitation input determines the accuracy of model simulation results. Radar estimator can provide a practical data source for hydrological modeling at a basin scale where the raingauge network is sparse. The result has important significance to improve the efficiency of hydrological modeling in large basins with scarce raingauges.

This research was supported by the Natural Science Foundation of China (40730633), the Major State Basic Research Development Program of MOST (2009CB825105), the Project of State Key Basic R & D Program of China (973 Program, 2010CB951002). The work has been published on Earth and Environmental Science, 2011, 25(11): 2669-2681. It can be linked from: http://www.springerlink.com/content/g62m467v98652x13/.

Land use and Soil distribution in the Manas watershed. (a) SWAT land use classification in the Manas watershed. RNGE, RNGB, PAST, FRST, BALD, WATR stand for Range-grasses, Rangebruth, Pasture, Forst-Mixed, Bald, Water respectively. (b) Soil distribution in Manas watershed.