基于高光谱的土壤重金属铜的反演研究  

Study on Heavy Metal Cu based on Hyperspectral Remote Sensing

作　　者：王维[1,2] 沈润平[1,2] 吉曹翔[2] 

机构地区：[1]南京信息工程大学气象灾害省部共建教育部重点实验室,江苏南京210044 [2]南京信息工程大学遥感学院,江苏南京210044

出　　处：《遥感技术与应用》2011年第3期348-354,共7页Remote Sensing Technology and Application

基　　金：国家973项目(G20000779);江苏省"青蓝工程";南京信息工程大学重点基金项目资助

摘　　要：为探讨高光谱遥感反演红壤重金属铜含量的可行性,研究采集了34个红壤性土壤样品,通过对350~2 500 nm波段范围光谱曲线进行测试和分析,建立了不同的土壤光谱变量与重金属铜含量多元回归关系模型,分析了土壤重金属铜与土壤化学组分以及土壤特征光谱的关系。结果表明,土壤重金属铜含量与土壤全铁和镁含量显著相关,而与土壤有机质的相关性不显著,表明红壤性土壤粘土矿物对土壤铜含量影响较大;与重金属铜含量相关性较好的波段在830 nm、1 000 nm和2 250 nm附近,且一阶微分模型精度（79%）高于反射率模型（66.26%）和倒数对数模型（67%）的精度。因此,一阶微分高光谱反演模型具有较好的快速估算土壤中重金属铜含量的潜力。In order to explore the feasibility of using Hyperspectral remote sensing technology to invert heavy metal Cu of red soil,based on the 350~2 500 nm band spectrum curve of 34 red loam samples.Through processing the spectrum curve and analysising spectral features,using three methods：spectral reflectance spectra,the first derivative and inverse-log,the stepwise regression equations are established.And based on statistical method,analysising the relationship of soil chemical components and soil spectral features with soil heavy metal Cu.The results show that the soil heavy metal Cu was well correlated with the soil Fe and Mg,while had non-significant correlation with soil organic matter.So claymineral had a significant effect on the content of Cu.The best relevance between soil spectra variables and soil heavy metal Cu appeared in band 830nm、1 000 nm and 2 250 nm.And precision of the model for first derivative（79%） was better than that of reflectivity（66.26%） and of Inverse-log（67%）.The hyperspectral model of soil heavy metal Cu had potential for quickly prediction of soil heavy metal Cu content.

关 键 词：高光谱 重金属铜 反射率 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] X833[自动化与计算机技术—控制科学与工程]