Soil heavy metal Pb concentration quantitative inversion method based on hyperspectral remote sensing

Abstract

Taking a mining area in the Liangshan section of the Anning River Basin as the research area, soil samples were collected, and the collected soil samples were respectively measured for Pb concentration and soil spectral reflectance, and the correlation between soil Pb concentration and soil spectral reflectance was analyzed, using the partial least squares regression method to construct the soil Pb concentration estimation model and carry out the accuracy test. The results show that: (1) The correlation between the original soil spectral curve and the measured value of soil Pb concentration is greatly improved after the first-order differential(FD), second-order differential(SD), reciprocal first-order differential(RTFD) and inverse second-order differential(RTSD) transformation, and there are many characteristic bands. (2) The best model for estimating soil Pb concentration is the model established by second-order differential, whose coefficient of determination (R2) is 0.56 and root mean square error (RMSE) is 1.28. It provides a reference for the retrieval of remote sensing hyperspectral soil Pb concentration in this area.