基于 Morzn＇s I 和 Kriging 插值在不同采样尺度下的土地利用程度分析--以重庆市石柱县为例

摘要

【目的】比较不同的采样方式和采样区的大小对土地利用程度空间分布和数据的精确性的影响，以期为土地利用程度空间分布现状和模拟提供合理的依据和技术支持。【方法】对比分析了重庆市石柱县基于 Moran＇s I 和Kriging 插值在不同采样尺度下的土地利用程度。【结果】各采样尺度下 Moran 指数均为正数，Kriging 插值结果表明不同的采样方式和采样区的大小对土地利用程度空间分布和数据的精确性存在一定的影响。【结论】①石柱县土地利用程度在各采样尺度下均存在正的全局空间自相关，全局空间自相关大小对采样方式和采样尺度较敏感。②采样区尺度的变化会影响土地利用程度值，同一种采样方式下，采样区尺度越小插值越精确，在采样区尺度较小的情况下，局部空间自相关能够精确定位区域土地利用程度聚集状况。%Abstrzct:[Objective]To compare the effect of different sampling methods and different sampling scales on spatial distribution of land utilization degree and accuracy of statistics.[Method]Com-paratively analyze the land utilization degree in Shizhu county,Chongqing municipality based on moran's I and Kriging according to different yardsticks of sampling.[Results]Moran's I index is positive under every different yardstick of sampling,and the results of Kriging interpolation show that different methods of sampling and different sampling scales have certain effect on spatial dis-tribution of land utilization degree and accuracy of statistics.[Conclusion]The conclusion in-cludes:First,the land utilization degree in Shizhu has positive overall space autocorrelation under every yardstick of sampling.Second,the change of yardstick has great effect on the data of land utilization,and under the same method of sampling,the smaller sampling area,the more accurate interpolation is.In a relatively small sampling area,autocorrelation in partial space can accurately locate the regional assembling situation of the land utilization degree.