基于辅助变量的森林半腐层厚度空间插值精度

Precision of spatial interpolation for forest duff layer depth based on secondary variable

基于地统计学方法,利用3种以海拔作为辅助变量的空间插值算法[局部平均的简单kriging法(simple kriging with locally varying mean,SKlm)、带有外部漂移的kriging法(krig-ing with an external drift,KED)和协kriging法(cokrging,COK)]计算了森林半腐层厚度的空间插值精度,并进行了交叉验证.结果表明:KED法既考虑了变量之间的空间变异,又考虑到影响局部空间变化的因素,与其他插值方法相比,其精度有很大提高;由于海拔与半腐层厚度之间的相关关系较弱,导致SKlm法的插值精度没有达到预期效果;COK法直接将海拔用于估计半腐层厚度,由于在边界地区缺乏采样点数据,因此边界地区的插值出现了多处突变区域.对比地统计学方法与距离反比权重法(inverse distance weighting,IDW)在本研究中的插值精度,除了KED方法的插值精度较高外,其余方法的插值精度均不及IDW,原因可能是利用辅助变量辅助地统计学插值时,主、辅助变量之间的相关关系在插值中起着重要作用.

Based on geostatistical method, three algorithms of spatial interpolation with elevation as a secondary variable, i. e. , simple kriging with varying local means （SKlm） , kriging with an external drift （KED）, and cokriging （COK）, were used to calculate the precision of spatial interpolation for the forest duff layer depth, and cross validation was conducted. The results showed that among the three algorithms, KED gave the highest precision because of its taking into account both the spatial variation among variables and the factors affecting local spatial change, SKlm did not yield expected precision because of the weaker correlation between elevation and forest duff layer depth, while COK directly used the variable elevation to estimate forest duff layer depth but many unexpected results yielded for the boundary area due to insufficient samplings. Comparing with the method of inverse distance weighting （IDW） , only KED had a higher precision of interpolation, while for SKlm and COK, their interpolation precision was lower, suggesting that when a secondary variable was used for geostatistical interpolation, the correlation between primary and secondary variables was of significance in increasing the precision of interpolation.