基于野外实测光谱的土壤汞含量估算研究

Estimation research on soil mercury based on field measured spectra

本研究以陕西省关中地区44个土壤样品的汞含量为数据源,使用ASD FieldSpec 4(350~2500nm)地物波谱仪获取土壤的反射光谱。比较不同含量的反射特征以及不同预处理方法对土壤重金属光谱反演模型建立的影响。对原始反射光谱进行多元散射校正、标准正态变换、标准化处理,结合反射率一阶微分、二阶微分、反射率倒数对数变换,筛选出不同数学变换形式下反射率与汞含量的敏感波段,采用偏最小二乘回归法建立高光谱估算模型。化学分析结果表明:该研究区域存在较为严重的Hg污染现象。模型结果表明:反射率随着汞含量的增加而减小,汞的敏感区域位于392~455nm、923~1040nm和1 806~1 969nm。原始反射光谱经过NOR、MSC、SNV变换后与Hg含量进行相关性分析后显著提高土壤中的Hg响应信息,模型的稳定性和预测能力显著增加。基于原始反射率倒数对数建立的偏最小二乘回归模型稳定性较好、精度较高,Rc^2=0.991 2,RMSEC=0.665;Rv^2=0.950 6,RMSEP=1.93,能够实现该区域汞含量的快速预测。

Taking the mercury contents of 44 soil samples in Guanzhong area of Shaanxi Province as the data source,the reflectance spectrum of the soil was obtained by ASD FieldSpec4(350~2500 nm)geophysical spectrometer.The effects of different levels of reflection characteristics and different pretreatment methods on the establishment of soil heavy metal spectral inversion models were compared.Multi-scattering correction,standard normal transformation and normalization processing of the original reflection spectrum,combined with reflectance first-order differential,second-order differential,and reflectance reciprocal logarithmic transformation,screen out the sensitive bands of reflectivity and mercury content under different mathematical transformation forms.The hyperspectral estimation model was established by partial least squares regression.The results of chemical analysis showed that there was a serious Hg pollution phenomenon in the study area.The model results show that:(1)The reflectivity decreases with the increase of mercury content,and the sensitive regions of mercury are located at 392~455 nm,923~1 040 nm and 1 806~1 969 nm.(2)After the correlation analysis of the original reflection spectrum with NOR,MSC and SNV and Hg content,the Hg response information in soil was significantly improved,and the stability and prediction ability of the model increased significantly.(3)The partial least squares regression model based on the inverse logarithm of the original reflectivity has good stability and high precision,Rc^2=0.991 2,RMSEC=0.665;Rv^2=0.950 6,RMSEP=1.93,which can realize the quick forecasting of mercury content in this region.