Estimating the Distribution of Heavy Metals in Soil from Airborne Hyperspectral Imagery Over Jilin Gongzhuling Gold Mining Area of China.

In this study, we used HyMap-C airborne hyperspectral imagery and ground samples collected synchronously to explore the estimation of soil heavy metal concentration. Preprocessing methods such as first-order derivative were used to enhance the weak spectral information related heavy metals. The multivariate stepwise regression method was used to select the spectral characteristics and establish the inversion model. The samples were divided into 3 parts, model set, validation set and test set. For the arsenic (As) the errors of the samples sets were 0.55, 0.75, 0.44, and the root-mean-square error were 51.20, 30.12, 32.78 mg/kg respectively. The results show that this method can predict the heavy metals arsenic in the study area.