【作者】 申柏希；

【导师】 朱建军； 戴吾蛟；

【作者基本信息】 中南大学 ， 地图学与地理信息系统， 2008， 硕士

【摘要】 随着工农业生产的发展,耕地土壤污染问题越来越突出,特别是耕地土壤重金属和农药等有机物污染尤为突出,对土壤环境各项污染因子进行评价、预测就显得尤为重要。本文以地统计学方法为基础,提出顾及地形等空间因素的协克吕格模型,使之更符合土壤环境实际情况,能够更加准确反应土壤污染程度的空间变异性,并应用南方某市的土壤监测数据,分别对内梅罗污染指数评价模型、克吕格模型和顾及空间因素的协克吕格模型的评价效益进行比较,得出顾及空间因素的协克吕格模型较优的结论。本文的具体工作及结论有:(1)以内梅罗指数为模型,对土壤重金属中Pb进行综合评价。根据该方法得出研究区域内各个子区的土壤污染等级,并对该结果进行评价。(2)以地统计学中普通克吕格方法做为土壤污染空间变异的统计模型,对土壤重金属中Pb在研究区域内进行综合评价,在研究区得到Pb的空间变异规律。并对该结果进行评价。结果表明:在大尺度下,由于受地形地貌等空间因素的影响,普通克吕格方法的误差较大,不能很好的反映土壤污染的空间变异性。(3)针对普通克吕格方法在大尺度上的不适应性,提出改进方法,即对研究区域先进行高程采样,得到DEM模型,再利用遥感图像对研究区域进行河流水系提取。最后根据上述等空间因素,对研究区进行分区,把大尺度的研究区最终分形为中尺度,小尺度,最终得到顾及空间因素的协克吕格模型。对土壤重金属中Pb在研究区域内进行综合评价,得到Pb的空间变异规律。(4)对上述三种评价模型的结果的精度进行评价。最终得出在研究区域为大尺度的情况下,顾及空间因素的协克吕格土壤环境评价模型是精度最高、预测能力最好的模型。更多还原

【Abstract】 According to the evaluation diffculties of soil pollution, Kriging interpolation is improved in this thesis. It makes it suitable to the reality of the soil environment, and can be able to more accurately reflect the extent of spatial variability of the soil pollution.This thesis takes some southern areas as an example, considers factors such as DEM, river, and the earth’s surface, and uses this method to interpolate. At the same time, it contrast the Neimeiluo pollution index model, the ordinary Kriging model and DEM Kriging model, then get a much better space count model. The main contents are:(1) Nemero index for the model of heavy metals was used in the soil Pb comprehensive evaluation. According to the study, we can obtain the soil pollution levels of the various sub-areas, and the results were evaluated.(2) Kriging method was used to evaluate the spatial variability of the Pb in the study areas. From the results, it can be concluded as following: in the large scale, because of the topography, and other spatial factors, ordinary Kriging methods can not well reflected the spatial variability of the soil pollution.(3) Because of the inapplicability of the ordinary Kriging, an improved method was developed in this thesis. The regional elevation sampling and remote sensing images ware studied to form DEM model and extract river line. Finally, the study area was divided into smaller areas according those facts. Then the modle based on Cokringing and spatial facts was used to get Pb on the comprehensive evaluation of spatial variability.(4) Compareing the accuracy of the evaluate results of the three models at last. Throught comparison, the third method is the best.更多还原