摘要
如何利用有限的样本数据来获得更为准确的土壤属性空间分布信息是土壤学研究的热点问题之一。利用福建省龙海市采集的1 133个耕地土壤样品,设计了结合地貌类型、土壤类型和土地利用类型等信息的5种克里格插值模型,研究县级尺度上土壤有机碳空间预测优化插值模型及其与样点密度的关系。结果表明:设计的5种插值模型预测精度均高于普通克里格法,但不同样点密度对插值结果影响较大。按0.5km×0.5km及以上的格网密度进行样点布设,采用土地利用现状类型结合土壤类型信息的普通克里格法(KDLTR)插值结果误差较小;按2km×2km的格网密度布设调查样点时,采用土壤类型信息的普通克里格法(KTR)插值结果误差较小;当格网大于4km×4km时,由于样点数少,各种模型的结果相差不大,可直接采用普通克里格法(KYJZ)进行插值。
How to use limited soil samplings to interpolation an accurate spatial distribution of soil properties is a hot issue in soil science.In order to study the optimized interpolation models and its relationship with soil sampling density,the research collected 1 133 soil samplings from Longhai County,Fujian Province,and designed 5 interpolate models combined with landform types,soil types,and land use types,etc.Results showed that the accuracy of 5 interpolate models are higher than Ordinary Kriging（OK） method.However,the size of soil sampling grid can affect interpolate results.Using interpolate model integrated with land use types and soil types can get more accuracy results when soil sampling collected from 0.5 km×0.5 km grids.When taking sampling form 2 km×2 km grids,the model integrated with soil types can get more accuracy results.But there are no variances between different models when sampling grid reaches 4 km×4 km,so it can use OK method to interpolate directly.
出处
《水土保持研究》
CSCD
北大核心
2011年第6期1-5,共5页
Research of Soil and Water Conservation
基金
福建省教育厅基金项目(JA11096)
