Runoff Simulation In Erhai Watershed Based On SWAT Model

The national and agricultural economy grew rapidly in Erhai Watershed in recent years. As the important supplier for water source, Erhai has been playing a vitial role in supporting the economic development. However, due to excessive fertilization, irrational fertilization structure and direct discharge into the rivers of farmland runoff and so on, Erhai faced increasingly severe eutrophication. In addition, climate change would lead to the variation of precipitation, evapotranspiration and nutrient in the whole hydrological cycle, and the water balance will be changed. Understanding the impact of climate change on the water resource is of importance for present and future water resources planning and management in Erhai watershed, and it is also the basic work to improve the watershed function and ecological enviroment. Therefore, study on the appropriate hydrological model for the watershed is of important implication for forecasting the hydrological process, learning the runoff feature at present and under future climate changes as well as resolving the water pollution problem.Appliing the distributed hydrological model SWAT (Soil and Water Assessment Tool), the study divided the watershed into 85 sub-watersheds and 943 hydrology response units (HRUs), and calibrated and validated actual monthly and daily runoff data from 2004 to 2008 in Liancheng hydrologic station. Results showed that: satisfactory monthly simulation was obtained with 0.72 for determination coefficient and 0.66 for Nash-Sutcliffe coefficient in the calibration period (in 2004-2007). The monthly and daily simulation was 0.88 and 0.7 for R2, 0.87 and 0.68 for Ens in the validation period (in 2008). The relative errors of monthly and daily simulation were very small. Overall, the correlation coefficient was above 0.7, and the model efficiency coefficient Ens was above 0.6, indicating a better and acceptable simulated runoff. Therefore, SWAT model is suitable for use in Erhai Watershed.Based on the runoff simulation in Erhai watershed in 1989-2008, the study analysed the space distribution of the average annual water quantity in each sub-watershed. The results showed that the water quantity was the least in some areas of Yousuo town, Sanying town, Niujie town and Dali town. It was the largest in some areas of Yuhu town, and followed by some areas of Shuanglang town, Wanqiao town, Yinqiao town and so on. On the basis of calibration and validation for SWAT model, we further studied variation in runoff under 24 combinations of precipitation and temperature. It was shown that the average annual runoff increased as precipitation increaes, and decreased as temperature rises. The average annual runoff was very sensitive to the change in precipitation, but it was less sensitive to the change in temperature. Assuming water used for industry, agriculture and inhabitant daily life was stable within short period of time, if temperature was invariable, the average annual runoff increased linearly with precipitation; and if precipitation was invariable, it decreased nonlinearly with temperature. The change in average annual runoff was significantunder different climate scenarios, and runoff increased the most under A5, while runoff reduced the most under A21. These two scenarios are the most advantageous and disadvantageous ones respectively in Erhai watershed.