Land Use Optimization In The Huaihe River Basin Based On Hydrological Responses

Soil erosion is still one of the major eco-environmental problems in China,which endangers the survival of human being and hinders the sustainable development of economy and society.Among the many factors causing soil erosion,land use is considered to be dominant.Therefore,reasonable land use planning is one of the most direct and effective methods to control soil and water loss.Traditionally,the evaluation of land use planning schemes in watershed is usually based on scenario analysis,i.e.,setting up multiple land use planning scenarios according to the characteristics of watershed,and then selecting relevant indicators for the analysis,so as to select the best plan among many.However,such scenario-based method is subjective and the number of land use planning scenarios is limited.The specific land use quantity and spatial adjustment scheme can not be given from the perspective of optimization.As one of the key regions for soil and water conservation,the Xixian watershed located in the upstream region of the Huaihe River basin was selected as the study area in this paper.The hydrological responses simulated by the distributed hydrological model were incorporated into the land use optimization model,and the land use pattern was optimized from coarse to fine in space by a progressive way.The main research contents and results were summarized as follows:(1)The SWAT(Soil and Water Assessment Tool)model was constructed in the study area.The model parameters were calibrated and validated with runoff and sediment data at the watershed outlet Xixian station.The results show that the Nash coefficients of monthly runoff simulation and validation period are above 0.80,and that of monthly sediment simulation and validation period are above 0.70,which indicates that SWAT model is capable of simulating runoff and sediment in the Xixian watershed.(2)Using the Grain for Green Project(GGP)as the land use planning operation,the scenario module of land use planning was developed.This module modified the input file of SWAT model and coupled SWAT executable program to perform hydrological simulation.Firstly,based on this module,the relationship among erosion control,dought risk and food security under two strategies,i.e.GGP based on agriculture land area and GGP based on Critical Source Areas(CSAs),was studied.The results show that the GGP will not cause drought crisis in Xixian watershed,and the GGP based on CSAs can better guarantee food security while controlling sediment.Then,1000 samples were randomly selected from all possible sub-basin combinations of GGP.The correspongding sediment yield results were simulated by the use of the scenario module of land use planning.BP neural network was then trained and validated accordingly.The results show that training period,validation period and prediction period are of high accuracy.It shows that the BP neural network model can accurately represent the complex input-output mapping relationship between land ues planning scenarios and hydrological responses.(3)A multi-objective integer programming model was constructed with the implementation of GGP at the sub-basin level as an independent variable.The dynamic relationships between land use change and hydrological responses simulated by BP neural network model were taken as the ecological objective,the area of converted agriculture land was used as the social objective,and the economic objective was presented by the decrease of GDP value.The basic genetic algorithm was used to solve the problem under the constraint of minimum per capita agriculture land area.According to the different emphases on the objectives,the GGP operations at sub-basin level with ecological target,social target and economical target were optimized,respectively.The results show that sub-basins optimized by those three schemes are concentrated in the upper reaches of Xixian watershed.The results also show that all optimal schemes perform better than the scheme of GGP based on CSAs in coordinating the contradiction between erosion control and food security.(4)According to the basic principle of evaluation factors selection and the availability of data,nine factors including DEM,slope,physiognomy,soil organic matter content,soil drainage condition,soil PH,annual average rainfall,accumulated temperature at 10℃ and distance from water source were selected.The weights of factors were determined by Analytic Hierarchy Process(AHP),and the suitability of agriculture land in the study area was evaluated comprehensively.It reveals that the highly suitable(S1),moderately suitable(S2),marginally suitable(S3)and unsuitable(N)agriculture land area accounted for 20.52%,48.42%,20.12% and 10.94% of the total area of the basin,and suitability levels are transitional in space,changing from the unsuitability in the source mountain region to the high suitability in the plain areas in the middle and lower reaches of the main stream.(5)According to the principle of overall planning and step-by-step implementation of GGP,further optimization of the spatial distribution of land use units in sub-basins is proposed.A multiple-phase GGP plan was formulatd under the constraints of the ecological target optimization results.A two-dimensional spatial coding genetic algorithm was used to optimize the spatial distribution of GGP at the grid scale in each stage.The integrated fitness function includes the evaluation results of agriculture land suitability,the spatial coordination degree of land use system and the sub-basin priority of GGP simulated by the SWAT model.The optimization results show that under the different stages,the optimized integrated fitness function values have been improved,ranging from 5.88% to 16.18%.In summary,the optimization method proposed in this paper can clearly define the specific areas of the implementation of GGP from sub-basin level to grid level,and provide an objective and scientific decision-making basis for water and soil conservation planning and management in Huaihe River basin.