摘要
从数据的统计特征分析、异常值筛选及处理、实验半变异函数计算及拟合等方面对普通克里格计算过程中的精度控制问题进行了初步研究,并以天津市东南郊区土壤砷(As)空间分布评估为例,分析了在多步骤精度控制下普通克里格法估计土壤污染物空间分布的可靠性。结果表明,研究区土壤As表现出了明显的累积特征,且监测值为正态分布,异常值较少且为局部异常值,实验半变异函数表现出明显的各向异性。在较严格的精度控制下,普通克里格估值的最优无偏特性可以得到较好的体现,其对土壤As的空间分布的估计精度可以达到较高的水平。
In this paper, the statistical characteristics analysis, outliers screening and treatment, experimental semi-variogram calculation and fitting were studied preliminarily as the main calculation process to improve the precision of ordinary kriging interpolation. A case of assessing the reliability of using ordinary kriging under multi-step precision controls to evaluate the soil arsenic （As） pollution in Tianjin southeast sub- urb was studied. The results showed that the soil As in the study area was marked accumulation, and the monitoring values were normal dis- tribution. There were few outliers in the monitoring values and they were all local anomaly values. The experimental semivariogram showed anisotropy. The optimal and unbiased characteristics of ordinary kriging had been brought into full play under the precision control and the estimation of spatial distribution of soil As could reach a higher level.
出处
《农业环境科学学报》
CAS
CSCD
北大核心
2009年第5期871-876,共6页
Journal of Agro-Environment Science
关键词
普通克里格
土壤
砷
精度控制
ordinary kriging
soil
arsenic
precision control
