Comprehensive and scientific control of forest stock is of great significance in maintaining the ecologi-cal balance.Based on 240 Angle gauge control sample data and Landsat8 images,the optimal model of the storage volume,remote sensing auxiliary variables and cross-variance function of the two were fitted respectively.A sam-ple independent spatial sampling method was designed using the spatial variation scale of the storage volume as the grid spacing.It was compared with three traditional sampling methods(simple random sampling,systematic sampling and stratified sampling).The spatial variation scale of forest stock is 7200 m.Beyond the range of 7200 m,the two points are independent of each other and no longer have spatial correlation;the variation scale of remote sensing auxiliary variables is 6600 m,and the variation scale of cross variation function of stock and re-mote sensing auxiliary variables is 6900 m,which is close to the variation scale of forest stock.Under the condi-tion of ensuring the sampling accuracy,compared with the three traditional sampling methods,the spatial sampling can only take 129 samples to make the sampling accuracy reach 89.46％,and the number of samples is far less than the number of the traditional sampling method(246).In the absence of sufficient sample data analysis of for-est stock variation function structure,a small amount of measured sample analysis of the spatial variation scale,spatial sampling method based on variation scale effectively reduce the sample information redundancy.Sample in spatial layout design is more efficient,reasonable and more representative.The results can provide a reference for the sampling survey of forest stock design.
NETL
NSTL
万方数据
