论文部分内容阅读
本文以土壤CEC和土壤全氮为研究对象,在河南省黄河以北6个地市选取870个样点,并随机均分为两个数据集,进行了制图对比分析;同时研究了Kriging法、IDW法和以点代面法制图结果的稳定性及其与精度的关系,以及影响因素。结果表明:采用实测数据与预测数据的交叉验证并不能用来衡量制图结果的稳定性,验证精度高低并不表示采用不同数据集制图结果可能会重现的概率;而且采用不同的方式进行精度检验,结论也会有所不同。采用不同抽样集合制图,Kriging法和IDW的结果较稳定,两者较接近,其相对误差>0.3的区域均不超过20%,且相对差异度高值区的空间分布格局相对分散;而以点代面法制图结果不稳定,相对误差>0.3的区域达到54.21%,且差异度较高的图斑相对集中,呈大片状分布。制图结果稳定性受到实测数据分布特征和局部地区土壤的高度变异性的影响,其中Kriging方法制图结果的不稳定性受样点分布格局的影响较另两种方法要大,而IDW法和以点代面法受实测数据自身变异性影响更明显。
In this paper, soil CEC and soil total nitrogen were taken as research objects, 870 samples were selected from 6 cities in the north of Yellow River in Henan Province and randomly divided into two datasets. The comparative analysis of cartography was carried out. Meanwhile Kriging, The stability of IDW and point-by-point method and its relationship with accuracy, as well as the influencing factors. The results show that the cross validation of the measured data and the predicted data can not be used to measure the stability of the mapping results. The accuracy of the verification does not mean that the probability of reproducing the results may be different with different data sets. Moreover, , The conclusion will be different. The results of Kriging method and IDW are more stable, the two are close, the relative error> 0.3 are all less than 20%, and the spatial distribution pattern of high relative value difference area is relatively scattered. The result of the surrogate method was unstable. The area with the relative error of> 0.3 reached 54.21%, and the spots with high degree of difference were relatively concentrated with a large patch distribution. The stability of the mapping results is affected by the distribution of the measured data and the height variability of the soil in the local area. The instability of the Kriging method mapping results is more affected by the distribution pattern of sampling points than the other two methods. However, the IDW method and the point Surrogate method by the measured data of the variability of their own more obvious.