用高程辅助提高土壤属性的空间预测精度

 【目的】探讨土壤属性变量与高程之间在何种条件下，可利用高程变量来辅助提高土壤变量的预测精度。【方法】用两种将高程作为辅助变量的克里格插值方法（协克里格法和简单克里格加变化局部平均值法）与没有考虑高程的普通克里格插值方法进行对比分析，用均方根预测误差和预测精度的相对提高值作为标准对3种方法的预测结果进行评价。【结果】对于交换性钾和pH值，协克里格法获得最精确的预测；对于Olsen-P、土壤有机质和有效锌，简单克里格加变化局部平均值法得到最精确的预测；而有效铜、有效铁和有效锰的最精确的预测结果则由普通克里格法产生。【结论】高程数据能够用来提高土壤特征的空间预测精度，但并不是对所有的土壤属性都适合；在利用高程数据来提高土壤属性空间预测之前，应该先对高程和土壤特征变量之间的线性相关关系、结构相关关系和全局趋势等进行仔细地分析，然后再选择适宜的方法。

Enhancing Spatial Prediction of Soil Properties Using Elevation

Abstract： 【Objective】This paper is concerned with increasing the accuracy of the spatial predictions of soil available nutrients (Cu, Fe, Mn, K, P and Zn), pH and soil organic matter (SOM) using elevation as ancillary variable. 【Method】The techniques applied were ordinary kriging (OK), cokriging (CK) and Simple kriging with varying local means (SKlm). The root of mean square error of prediction was used as the comparison criterion to assess the performance of each prediction method. 【Result】CK resulted in the most accurate estimates for K and pH, and SKlm for P, SOM and Zn, whereas both bivariate interpolation methods could not improve the accuracy of prediction for Cu, Fe and Mn relative to OK. 【Conclusion】The results suggest that the factors be responsible for prediction methods should be examined carefully before deciding on the most appropriate method of prediction