基于土壤重金属含量指数变化的冠层光谱遥感分析

Canopy Spectral Remote Sensing Analysis Based on the Change of Soil Heavy Metal Content Index

为提高水稻冠层光谱反演水稻田土壤重金属含量的精度,针对实测土壤重金属含量进行信息提取研究,对其分别进行了对数变换、最大值变换和Box-cox变换。在实验分析中,为得到土壤重金属(Cu、Pb、Cd)反演模型,将土壤重金属及其变换数据与冠层光谱及其变换后的数据进行偏最小二乘回归分析,并分析所得模型精度和稳定性。分析结果表明,对测定重金属数据进行变换能够普遍提高模型的精度与稳定性;对属重度污染重金属Pb和中度污染重金属Cd有较好的预测结果,而对未形成污染的Cu则难以形成有效的预测分析。其成果对耕地土壤重金属含量监测评价具有重要的参考价值。

In order to improve the accuracy of rice canopy spectral inversion of soil heavy metals,this study carried out the information extraction of the measured soil heavy metal content,logarithmic transformation,maximum transformation and Box-cox transformation respectively.In the experimental analysis,the partial least squares regression analysis was performed with canopy spectra and transformation data to study the soil heavy metals(Cu,Pb and Cd)inversion models,the accuracy of the model and the level of pollution in the study area were analyzed.The results showed that the accuracy and stability of the model could be generally improved by the changing the measured heavy metal data.In the prediction of heavy metals,it is found that there was a good prediction result for heavy metal Pb and moderately polluted heavy metal Cd in the study area,while it was difficult to form effective analysis for non-polluted Cu.The results are the important reference for monitoring and evaluation of soil heavy metal content in cultivated land.