Study On Eco-hydrological Effects In Huangshui River Basin Based On SWAT Model

In recent years, the study of watershed eco-hydrological effects on climate andland use/cover changes has been a focus. Watershed eco-hydrological effects are thecombined results of climate and land use/cover change. Climate and land use/covercan cause great changes of underlying surface, the runoff mechanism andhydrological regime. As the responses of simulated runoff and sediment to differentwater balance calculations on climate and land use/cover change are not the same, itpossesses a very important theoretical significance and practical value to study thehydrological effects of climate and land use/cover changes in different regions.In this paper, Xining section of Huangshui River basin is selected as the studyarea and many factors such as climate, land use/cover, terrain, soil and etc. are takeninto account. First of all, the changes of precipitation, temperature, runoff, sedimentbetween1955and2010，and land use/cover of three periods are analyzed. Then basedon RS and GIS technology, the runoff and sediment between1990and2010inupstream area of Xining station are simulated with SWAT model. The main contentsand results are as follows:(1) With linear regression, moving average, Mann-Kendall test and slidingT_test, the changes of meteorological and hydrological factors are analyzed. It hasbeen shown that precipitation and temperature increased significantly while runoffand sediment dropped significantly between1955and2010, there was only amutation of runoff in1967.(2) On the basis of RS and GIS technology, the changes of land use/cover in1990,2000and2010are analyzed. It has been shown that the transformations of landuse/cover between1990and2010mainly take on the change from grassland toforestland and farmland as well as from farmland and grassland to building land,meanwhile the transformations between2000and2010are bigger than in the first tenyears. (3)The spatial database and SWAT model are established on the basis ofanalyzing the changes of meteorological and hydrological factors as well as landuse/cover in study area. The parameter sensitivity analysis, calibration andverification of runoff and sediment are separately executed with SWAT_CUP underthe cases that measured runoff and reductive runoff are served as observations. Theresults show that the calibration and verification of monthly runoff under both caseshave reached the standard that NS coefficient is more than0.6and correlationcoefficient more than0.5, and the simulated effects of runoff under the reductiverunoff are better, thus SWAT model can well simulate the watershed runoff. Due tothe influence of sandpits in upstream, the simulations of sediment under both casesare largely different from the actual.(4) In the uncertainty analysis of model, different watershed division numbersand land use/cover are considered under the reductive runoff. It has been shown thatthe impacts on watershed runoff are both very small.(5) Different land use/cover and climate scenarios are assumed on the basis ofanalyzing climate and land use/cover changes in watershed. With comparisons of thesimulated runoff under different scenarios, it has been found that runoff will rise withthe increase of forest or grass coverage, and the forestland has a greater impact onrunoff than the other land use types in the same areas; the increase of precipitationwill lead to the rise of runoff, temperature is just on the contrary, and the impact ofprecipitation on runoff is bigger than temperature. Through the comparisons ofsimulated runoff under different assorted scenarios, it has been found that the runoffwill rise with the increase of vegetation coverage on the same climate.