遥感和GIS辅助下流域养分迁移过程的计算机模拟

Distributed modeling of nutrient transport in basin with support of remote sensing and geography information system

农业非点源污染问题已成为我国重要的环境污染类型之一 .利用空间模型对流域N、P的迁移过程进行计算机模拟 ,是研究非点源污染控制方法的一个有效手段 .设计了遥感和GIS技术辅助下的流域水分和养分迁移过程分布式的模拟方法 ,包括模型选择、流域的空间离散化和参数化、模型模拟和结果验证 3个步骤 ,为控制流域水肥流失提供了思路 .以江西兴国潋水河流域 (5 79km2 )为研究区域 ,选择美国农业部设计的SWAT模型 ,设计了流域 子流域 水文响应单元的空间离散方案和实现步骤 .首先依据地形特征 ,将整个流域分割成多个子流域 ,每个子流域内部通过叠加统计分析 ,生成单一土地利用类型和土壤类型组合的水文响应单元 .土地利用参数用TM遥感图像进行监督分类获得 ,土壤参数化利用地统计学采样和插值分析获得 .对 1991～ 2 0 0 0年的初步预测结果表明 ,SWAT模型能够较好地模拟潋水河流域的径流水量和泥沙的变化 ,产水和产沙 10年平均预测精度分别为 89.9%和 70 .2 % .

Agricultural non-point source pollution has become serious in our country. Modeling the processes of nutrient(especially nitrogen and phosphorus) transport in basin and evaluating the adopted management practices are important for controlling the impact of non-point pollution on environment. A research scheme for distributed simulation of nutrient transport processes in Lianshui Basin, Xingguo County, Jiangxi Province was designed, with the support of remote sensing (RS) and geography information system (GIS). The research procedure included model selection, discretization and spatial parameterization of the basin, prediction, and validation. SWAT model was selected, and basin-subbasin-hydrological response unit discretization scheme was designed. Supported by RS and GIS and based on the topographic features of the watershed, the subwatershed discretization divided the watershed into subbasins, and each subbasin could be further partitioned into multiple hydrologic response units (HRUs), which were unique soil/land use combinations within the subbasin and modeled through statistical spatial overlay analysis. The parameters of land use were obtained from the supervised classification of TM imagery based on field training samples, and those of soil were obtained from field sampling and further interpolated through geostatistical method. The simulation was carried out by using the data from 1991 to 2000. The results showed that the simulation accuracy of annual runoff water yield and sediment yield was 89.9% and 70.2%, respectively.