Hydrological Responses To LUCC Based On Swat Model In Baimahe Basin

As the lodgement and penetration point of global environmental changes sciences,Land Use and Land Cover Changes (LUCC) are not only the final embodiment ofhuman-environment interaction, but the most obvious landscape patterns of changesof the earth surface system. And research of hydrological responses to LUCC is one ofthe key issues in global changes sciences and hydrologic sciences. In long terms,climate changes are the main reasons influencing basin hydrological processes, whilein short terms, LUCC driven by population increasement, economic improvement,urbanization and the policy of Reform and Openness will no doubt affect basinhydrological precesses and water balance, and lead to ecological problems such aswater and soil loss and land degradations. Thus, research of hydrological responses toLUCC has great meanings in exploration and management of water resources andensuring their sustainable development.From 1970s, distributed hydrological models have been widely used insimulating basin hydrological processes. Among these models, SWAT model isinvented by US Department of Agriculture (USDA)–Agricultural Research Servicein 1994. It is based on the theory of water balance and has the advantages ofconvinency, good application, long-term persistent simulation, degredation of spatialdifferences, high simulation efficiency, and so on. Therefore, SWAT model is widelyused in research of hydrological responses to LUCC.Baimahe basin, as one important sub-basin of Nansihu basin in Shandongprovince, is chosen here as the research region. Based on 3S technologies, TM imgesof year 1987, 1997 and 2007 are applied to obtain three LUCC patterns in more thantwenty years. By using CA-Markov model, LUCC pattern in 2012 is predicted. Then,SWAT model of the basin is constructed to simulate and calculate the contributioncoefficients of different lan use types on runoff; and after finding the mutation point of climate changes and runoff changes, SWAT model is established to simulate andcalculate the coefficients of climate and LUCC on runoff.The mian conclusions of this thesis are listed as follows:(1) From 1987 to 2007, farmland, construction land, forest and shrub land arefour main land use types, occupying 96% of the total area of the basin and leavinggrass land, waters and unused land subdominant ones. Elevation, gradient and ascentcan all affect LUCC patterns in some degree. And farm land is distributed in largerareas with the growth of elevation and gradient, so is construction land, but it is lessaffected by ascent. Froest and shrub land are two main land use types above 250m.Compared with the above three land use types, grass land, waters and unused land areunobviously affected by three topography factors.(2) Results of CA-Markov model show that, in 2017, the main land use types arestill farmland, construction land, shrub land and forest, and from 2007 to 2017, areasof construction land and shrub land increase, while areas of farmland and forest keepon diminishing.(3) Using ArcGIS9.3, SWAT model is established. Selecting actual measurementrunoff data from 1991 to 1995 and using SWATCUP, the established SWAT model iscorrected to achieve the best simulation result. And actual management data in twoperiods of 1986-1990 and 1996-2000 is selected to validate the corrected SWATmodel. Result shows that SWAT model is applicable in Baimahe basin.(4) Keep climate data input from 1986 to 2000 and late runoff of different LUCCpatterns respectively. Result shows that with changes of LUCC patterns in 20 years,runoff diminishes gradually. In order to embody the effects of LUCC on runoff moreobviously, 4 imaginable LUCC patterns are established.They are 2017 predictionpattern (F1), farm land conversion to forest pattern (F2), severe land degredation (F3)and forest construction pattern (F4). Simulation result shows that compared withLUCC pattern in 2007, from F1 to F4, runoff keeps on increasing. Muti-objectivedecision-making function is established to calculate contribution coefficients of thefour main land use types on runoff. Results show that forest, shrub land andconstruction land can all facilitate production of runoff, while farm land can prohibite production of runoff.(5) Different SWAT models are established according to different combinationsclimate changes and LUCC pattern. And contribution coefficients of climate changesand LUCC patterns on runoff are simulated and calculated to be 42.85% and 56.97%respectively, showing that climate changes and LUCC pattern are two main factorsinfluencing runoff changes of Baimahe basin, and between the two, the latter is themore influencing one.