ArcGIS支持下的样本稀疏山区空间插值模拟探讨

样本稀疏地区空间插值法对区域化变量的精准管理具有重要意义。基于ArcGIS 90，在分析土壤属性空间分布特征的基础上，提出并构建了基于不同土壤类型的土壤特性空间预测模拟模型，对比了传统方法与改进方法空间插值精度，实现了数值插值在复杂地理环境区域的应用，得到以下结论：(1)基于经度、纬度、海拔高度及坡度等地理因子的土壤基础环境因子的空间预测模拟模型，突破以往只能描述土壤属性在水平方向变化的局限，较客观、合理地反映土壤属性随地理位置及海拔高度的立体变化特征；(2)基于不同土壤类型回归模型来增加样本点以推断评价指标在无取样地区的分布状况的处理方式具有一定的数学理论支撑，有效降低了插值误差，提高了评价精度，使评价结果更加接近现实。

SPATIAL INTERPOLATION RESEARCH IN HILLY REGIONS WITH SPARSE SAMPLES BASED ON ARCGIS

A study on distribution of soil nutrient on the field scale is important for improving agricultural management,and for assessing effects of agriculture on environmental quality and degrees of the influence of some random factors.However,soil nutrients are highly heterogeneous whether on a large scale or a small scale,and their heterogeneity results from many processes acting and interacting across a continuum of spatial and temporal scales.In this paper,a new method setting suppositional samples over region...