摘要
选取江西省乐安河及其支流流域作为研究区域,探讨了使用土壤高光谱数据间接反演其重金属(Cu、Zn、Pb)含量的方法。选用偏最小二乘模型对土壤有机质含量进行高光谱反演,引入人工神经网络回归模型建立土壤有机质含量与重金属含量的相关关系,从而提取出土壤中痕量级的重金属元素,并针对其空间分布情况进行关联分析和对比分析。实验结果表明,此方法在反演Cu、Zn元素时可有效反映其空间分布特征,具有在类似泛滥平原区域推广的适宜性,也为该区域土壤及水文生态环境监测提供了相关参考。
At present, supervision on the heavy metal contents in soil attracts increasing attention in the field of geography. This paper selects Lean River and its tributary basin in Jiangxi Province as the study area,discusses an inversion analysis method for the heavy met- als (Cu, Zn, Pb) content in soil by using the hyperspectral data. According to the physical and chemical properties of the basin soil, the PLS model is used for organic matter content inversion based on the hyperspectral data; and then, the artificial neural network model has been used to set up the correlation between the organic matter content and the heavy metal content. Finally, the spatial distribution char- acteristics of the heavy metals content in soil are summarized by using correlation analysis and comparative analysis methods. The experi- mental results provide a powerful reference for the ecological environment monitoring of the basin soil, and strongly show that this method is satisfactory, and is also suitable for promotion in similar flood plain regions.
出处
《地理与地理信息科学》
CSCD
北大核心
2015年第3期26-31,F0002,共7页
Geography and Geo-Information Science
基金
国家自然科学基金项目(41301377)
关键词
土壤重金属含量
高光谱反演
偏最小二乘回归(PLS)
人工神经网络
heavy metals content in basin soil
retrieval with hyperspectral data
partial least squares regression (PLS)
artificial neural network