洱海流域用地变化的智能体建模与非点源污染响应模拟  

Agent-based modeling of land use change and simulation of response on nonpoint source pollution for Erhai Lake watershed

作　　者：许泉立[1,2,3] 杨昆[1,4,3] 李建[4,3] 杨玉莲[4,3] 

机构地区：[1]武汉大学遥感信息工程学院,湖北武汉430079 [2]云南师范大学旅游与地理科学学院,云南昆明650500 [3]西部资源环境GIS技术教育部工程研究中心,云南昆明650500 [4]云南师范大学信息学院,云南昆明650500

出　　处：《水利学报》2014年第11期1272-1283,共12页Journal of Hydraulic Engineering

基　　金：国家高技术研究发展计划主题项目(2012AA121400);国家自然科学基金资助项目(41461038);高等学校博士学科点专项联合基金(20115303110002);云南省科技厅自然科学基金项目(2011XX2005)

摘　　要：洱海流域水环境污染具有显著的人类活动-土地利用/覆被变化-非点源污染链式驱动效应,有必要分析其驱动机制,模拟和预测流域的用地变化及其非点源污染响应过程。本文利用智能体模型模拟人地交互作用过程,预测洱海流域未来土地利用/覆被动态变化的结果,进而利用SWAT模型计算不同用地情景下流域主要非点源污染指标(氮和磷)的产出量和时空分布,最后,利用统计学分析方法与GIS技术,对综合模拟结果进行了精度验证、敏感性检验以及空间分析与可视化表达。结果表明:(1)流域过去和未来10年的土地利用/覆被变化表现为耕地的持续性减少(-19.28%)与建设用地的不断增长(35.27%);(2)氮和磷在时间与空间上具有相对一致性,说明流域土地利用/覆被变化对氮、磷具有相似的水环境效应;(3)未来10年流域非点源污染主要来自于农耕区,建议在保护耕地的同时,优化农耕区的种植结构和生产模式,以减少农业生产的非点源污染贡献率;(4)未来10年流域耕地面积持续性缩减,农业非点源污染呈现逐年降低趋势;同时随着建设用地不断增加,不透水表面的快速增长可能导致的城市非点源污染风险需要引起足够的重视。There is a significant chain-driven effect on water environmental pollution for the Erhai Lake Watershed, which is nonpoint source pollution due to land use and cover change （LUCC） caused by hu？man activities. Therefore,it is necessary to analyze this chain-driven,simulate and predict land use chang？es as well as response of nonpoint source pollution of the Erhai Lake watershed. An agent-based modeling （ABM） was used to simulate the interaction mechanism between human and land and to predict the results of LUCC; and then, based on the LUCC forecasting and SWAT model, the outputs and spatial-temporal distribution of major nonpoint source pollution loads were computed, such as nitrogen and phosphorus un？der different land use scenarios. Finally, combined with statistics methods and GIS technologies of spatial analysis and visualization, the processing of validation and sensibility verification of model are discussed, and comprehensive simulated results are analyzed. The results show that：（1） the major interaction between humans and the land over the next ten years will force a persistent reduction in farmland （-19.28 %） and a continuous increase in urban land （35.27 %）;（2） the relative consistency of nitrogen and phosphorus in time and space shows that LUCC has a closer water environmental effect on nitrogen and phosphorus in the study area;（3） the nonpoint source pollution of watershed in the next 10 years will mainly come from the agricultural area, the agricultural planting structure and pattern of production areas should be optimized to reduce nonpoint source pollution contribution rate of agricultural production while taking measures to pro？tect farmland;（4）in the next 10 years the farmland areas will reduce continually and agricultural nonpoint source pollution will appear to reduction trend year by year;meanwhile, as increasing land of construc？tion, a possible rapid increase in urban nonpoint source pollution caused by impervious surfaces needs to attract enough a

关 键 词：土地利用/覆被变化 流域水环境效应 人工神经网络 智能体建模 非点源污染响应 

分 类 号：S156[农业科学—土壤学]