| The development status of the application of SWAT (Soil and Water Assessment Tool) model on hydrological simulation and model construction, including the selections of the basic data and different simulation methods, the analysis of sensitivity parameters and model validation are introduced and analyzed. There are three aspects in the model application. Most of the study areas are on the semiarid inland. The selections of the basic data and subjective factors synthetically affect the simulation efficiency of the model. The research results lack comparability , which emphasize particularly on the simulation efficiency of the model affected by certain factor. Runoff is a synthetical effect of different hydrological processes. The total runoff yield simulation of the watershed is used to evaluated the applicability of the model in most of research results, which make the results lack reliability in the simulation of hydrological processes. The hydrological response to land use-cover changes and climate changes are simulated in the model application, but the simulation are mostly based on the virtual scenarios and factor simplification. So the influence factors should be synthetically analyzed in the research of model application to improve the simulation efficiency of the model.Comprehensive consideration to the impact of the mesh size of the digital elevation model from 30m to 1500m and the subwatershed partitioning for the distributed SWAT model, annual runoff and sediment yield in Jinjiang basin, which is located at coastal area of Fujian Province, southeast of China., were evaluated. The results show that DEM of 30m mesh size and the suitable threshold subwatershed size (2700hm2) based on the fractal dimension can be determined to optimize the input data preparation for SWAT.Annual water yield, monthly streamflow and baseflow of three gauging stations in Jinjiang basin were evaluated based on SWAT model. The model was calibrated with the measured runoff data of 1972—1975, and six sensitive parameters including CN2, SOLA WC, ESCO, etc. were determined. The simulation results were evaluated by Ens, MD , RE and R2. The model was further validated with the runoff of 1976 -1979, which suggested excellent results. It shows that the calibration and validation with three sites result in more realistic parameter values across both the hydrological processes and the geographic areas, and improve the simulation efficiency. SWAT model could be successfully used to model the runoff in the southeast coastal watershed. Better results can be obtained if the simulation with wet condition data (above average rainfall for the basin) than with dry condition data (below average rainfall). In addition, during the low-flow period, the reservoir operation greatly affects the simulation results.Based on the results of runoff yield simulation, the sediment yield from 1972 to 1979 of the upper watershed of Jinjiang basin is simulated with the distributed SWAT model. The simulation results are obtained with the accuracy of a R2 of 0.78 and an Ens of 0.79 over the 8 years continuous simulations on monthly sediment yield in Xixi River, which is the main source of sediment yield in Jinjiang basin. It shows that SWAT model could be successfully used to model the sediment yield in the southeast coastal watershed.The environment impact of vegetation restoration of degraded ecosystem includes hydrological process, soil erosion and so on. Based on the validation results of the yield of runoff and sediment, different revegetation scenarios were simulated with SWAT model in order to evaluate the impact of vegetation changes on hydrological effects in Xixi basin. It shows that SWAT model could be successfully used to model the hydrological process of revegetation of degraded ecosystem. The results demonstrate that revegetation has a great impact on the sediment yield. Compared with the grassland rehabilitation, the forest rehabilitation can reduce runoff and sediment yield effectively in the farmland of 15°～25°slope.
|
PDF Full Text Request