Optimization Of Regional Land-use Structures Based On Prediction And Management Of Non-point Source Pollution Under Conditions Of Uncertainty

Rapid socioeconomic development,leading to significant changes in land-use patterns,has influenced the output of regional nonpoint-source(NPS)pollution and posed threat to water body and human health.It is essential to predict regional land-use changes and evaluate NPS pollution loads under the influence of multiple socioeconomic factors.On this basis,it is crucial to optimize regional land-use patterns considering NPS pollution control for improving quality of water environment and maintaining sustainability of socioeconomic development.However,multiple uncertainties,such as interval and fuzzy information,exist in the processes of land-use changes and NPS pollution export.These uncertainties limit the effectiveness of the management of regional land-use patterns and control of NPS pollution.Thus,it is of great importance to conduct study on optimization of regional land-use patterns based on prediction and management of non-point source pollution under conditions of uncertainty.In this research,an integrated land-use prediction and optimization(ILUPO)model based on system dynamics(SD),export coefficient,the conversion of land-use and its effects at small region extent(CLUE-S),interval linear programming,and fuzzy parameter programming models was proposed.The ILUPO model can provide future land-use patterns and NPS pollution loads,and also help optimize the patterns under multiple scenarios of pollution reduction and decision-making tendency.Interval and fuzzy uncertainties in the processes of land-use changes and NPS pollution output can also be effectively addressed.The developed model was applied to the Xinfeng County,a region in the upper reaches of the Xinfengjiang Reservoir Watershed in Guangdong Province.The main works and results are shown as follows:(1)Regional land-use structures under different socioeconomic scenarios in 2030 were predicted in this study.Based on grey relational analysis,factor analysis,and linear regression model,an SD model for prediction of regional land-use change was developed.Input variables of the SD model under the low and high-speed development scenarios during the prediction period(2030)were obtained through grey prediction,linear regression,geometric mean,and arithmetic mean methods.Results show that under the high-speed development scenario,cropland area in 2030 would be approximately 7.5% smaller than that in 2020,while waterbody areas would increase by approximately 16.0%.In contrast,areas of these two types of land-use would show opposite variation trends under the low-speed development scenario.Construction land area would respectively decrease by approximately 33.2% and 30.0% under the low and high-speed development scenarios.On the contrary,forestland and grassland would show an increase trend under both scenarios.Specifically,forestland area would increase by approximately 0.2% and 0.1%,and grassland would increase by approximately7.0% and 13.2%,respectively.(2)Regional land-use patterns and the NPS pollution loads were analyzed based on the predicted land-use structures.Drivers of spatial variation in regional land-use,including socioeconomic and natural environment factors were collected.The predicted areas of different land-use types in the study area were allocated through CLUE-S model.The total amount and spatial distribution of NPS pollution loads were also estimated and simulated through export coefficient model and spatial analysis.Results show that the regional land use conversion would be more active in the high-speed scenario,in which the conversion frequency of cropland and forestland would be the highest.The total nitrogen(TN)and total phosphorus(TP)loads of regional land-use pattern in 2020 are 1237.9 and 46.9 t,respectively.Variation of the predicted land-use patterns would lead to an increase of TN loads in 2030 under the low and high-speed scenarios compared to those in 2020,while the TP loads would show relatively complex variation trends.Concurrently,the TN and TP pollution loads per unit area would increase significantly in the densely distributed areas of construction land and cropland in 2030.(3)An inexact optimization model was developed through integrating export coefficients,interval and fuzzy parameters programming models.The optimization model can effectively address interval and fuzzy uncertainties.Optimal land-use patterns under multiple scenarios of pollution reduction targets and decision-making tendencies can also be obtained through the developed model.The optimization results show that pollution loads in the study area cannot be reduced by >5% through land-use adjustment.Among the five land-use types studied(i.e.,cropland,forestland,grassland,waterbody area,and construction land),different scenarios of pollution reduction targets and decision-making tendencies would have greater influences on the optimized areas of the cropland,forestland,and construction land.The optimized cropland and forestland would still be the key sources of regional NPS pollution.The results under multiple scenarios of pollution reduction targets and decision-making tendencies can provide regional managers with multiple decision-making schemes.The proposed model has strong applicability in the optimization of regional land-use structure considering non-point source pollution control.It can provide methodological support for decision makers in non-point source pollution control and land-use structure adjustment.