土壤信息不确定性在空间分析中的传递是数字土壤评价中的关键问题。本文提出使用序贯高斯模拟（Sequential Gaussian simulation，SGS）和拉丁超立方抽样（Latin hypercube sampling，LHS）相结合的方法（即SGS-LHS），来应对该问题，目的在于充分利用SGS和LHS各自的优点，互补各自的缺点，以提高不确定性传递分析的准确性和效率。这种方法（包括两种途径：SGS-LHS1和SGS-LHS2）和SGS、LHS一起被应用于香港农田土壤质量评价中，并进行了比较。结果表明：（1）SGS-LHS分析所得的不确定性结果与SGS接近一致，与LHS则有一定的差别，但差别不大；（2）SGS-LHS估计不确定性的准确性与SGS接近一致，且两者均高于LHS，尽管LHS估计的置信度区间平均宽度略显精确。
Transmission of uncertainty of soil information in spatial analysis is an issue critical to digital soil assessment. To cope with the issue, coupling of the sequential Gaussian simulation (SGS) and the Latin hypercube sampling (LHS) methods, i.e., SGS-LHS, was proposed with a view to making full use of the advantages of the two methods as complementation to overcome their respective drawbacks, so as to improve accuracy and efficiency of the transmission and analysis of the uncertainty. The new method, (including two pathways: SGS-LHS1 and SGS-LHS2), SGS and LHS were tested in soil quality assessment of farmlands in Hong Kong for comparison. Results show that the uncertainty of the analysis using the SGS-LHS method was similar to that using SGS, but different to a certain, rather a big extent from that using LHS, and the combination method was approximate to SGS, but higher than LHS in accuracy, although LHS appeared to a little bit more accurate in terms of mean width of the confidential intervals. Therefore, the combined SGS-LHS method is recommended for analysis of transmission of soil information uncertainty in spatial analysis.
孙孝林,王会利,曹继钊.应用SGS和LHS分析数字土壤质量评价中的不确定性[J].土壤学报,2014,51(5):963-973. DOI:10.11766/trxb201309130415 Sun Xiaolin, Wang Huili, Cao Jizhao. Application of SGS and LHS to analyzing uncertainties in digital soil quality assessment[J]. Acta Pedologica Sinica,2014,51(5):963-973.复制