土壤重金属Cd污染指数的适宜插值方法和合理采样数量研究

发布时间：2018-05-14 11:53

本文选题：插值方法 + 采样数量　；参考：《土壤通报》2016年05期

【摘要】：对局部存在重金属污染地区采用适宜的插值方法和布设合理的采样点对重金属污染状况监测具有重要的意义。运用单因子指数法得到土壤重金属Cd污染指数,并在镇域内布设2033个采样点的基础上,通过随机抽样法抽取1830、1423、1017、610、203五个采样点样本子集。首先,运用普通克里金法(OK)、径向基函数法(RBF)和反距离权重法(IDW)对该地区土壤重金属Cd污染指数进行插值预测,并通过交叉验证法进行精度检验。然后,在反距离权重法的基础上,对五个采样子集进行插值精度分析,得到大致合理采样数量。结果表明:(1)利用全集2033个采样点对3种插值方法进行交叉验证分析可知,RMSE表现为IDW(3.018)RBF(2.942)OK(2.837),ME表现为OK(-0.0736)IDW(0.0214)RBF(0.0096),MAE表现为IDW(0.5668)RBF(0.5575)OK(0.5227),3种插值方法在整体预测精度上差异不明显。而对于污染区域的识别,IDW在轻度污染区、中度污染区和重度污染区预测上表现出较大的优势,能较好的反应污染区域的空间变异特征。因此,认为IDW为较适宜的空间插值方法。(2)对不同采样数量的样本进行交叉验证分析可知,RMSE、ME和MAE在1017个采样点到610个采样点误差变化幅度分别为29.84%、71.31%和36.99%,误差增加幅度较前三个子样本间明显增大。在空间特征识别方面,2033、1830、1423和1017个采样点反映的污染区的空间分布特征非常相似,610和203个采样点预测的污染区域面积明显扩大,对各级污染区域的空间特征细节表现能力较差。因此,对于该镇域内的土壤重金属Cd污染指数的研究,1017个左右采样点是比较合理的采样数量。
[Abstract]:It is of great significance to monitor the pollution status of heavy metals by using appropriate interpolation method and setting up reasonable sampling points for local heavy metal polluted areas. The CD pollution index of soil heavy metals was obtained by single factor index method. On the basis of setting 2033 sampling sites in the town area, the subsets of 1830 ~ 1423 ~ 1017610203 samples were selected by random sampling method. Firstly, the common Kriging method, radial basis function method (RBF) and inverse distance weight method (IDW) are used to predict the CD pollution index of soil in this area. Then, on the basis of the inverse distance weight method, the interpolation accuracy of five sampling subsets is analyzed, and the approximate reasonable sampling quantity is obtained. The results showed that the RMSE of IDW3.018 RBFU 2.942 OKO 2.837Me showed that IDW0.5668RBFU 0.5575OKO 0.52272727 had no significant difference in the overall prediction accuracy by using 2033 sampling points of the whole set. The results showed that the RMSE showed no significant difference in the overall prediction accuracy of IDW0.5668 RBFU 0.5575OKO 0.52272770.The RMSE showed no significant difference in the overall prediction accuracy of IDW0.5668RFU 0.5575OKU 0.522727. The IDW of the polluted area shows a great advantage in the prediction of the mild, moderate and heavy polluted areas, and can better reflect the spatial variation characteristics of the polluted areas. Therefore, It is concluded that IDW is a more suitable spatial interpolation method. (2) the cross validation analysis of samples with different sampling numbers shows that the error range of RMS Eime and MAE from 1 017 sampling points to 610 sampling points is 29.84% and 36.99% respectively, and the error increase range is higher than that before. The size of the three subsamples was significantly larger. In the aspect of spatial feature identification, the spatial distribution characteristics of polluted areas reflected by 2033N 18301423 and 1017 sampling sites are very similar to those of the pollution areas predicted by the sampling sites of No.610 and 203, and the spatial characteristics of the polluted areas at different levels of pollution are not well represented by the spatial characteristics of the contaminated areas. Therefore, for the study of CD pollution index of soil heavy metals in the town area, 1 017 sampling sites are more reasonable for sampling.
【作者单位】：中国地质大学土地科学技术学院;国土资源部土地整治重点实验室;北京师范大学减灾与应急管理研究院;江苏省地质调查研究院;
【基金】：国土资源部公益性行业科研专项课题(201511082-02)资助
【分类号】：X53;X833

【相似文献】