Qualitative analysis and mapping of heavy metals in an abandoned Au–Ag mine area using NIR spectroscopy

Reflectance spectroscopy covering the visible and near-infrared ranges (400–2,400 nm) is known as a rapid and nondestructive method in the characterization and quantitative analysis of many components of interest. In this study, the potential use of spectral absorption feature parameters (SAFPs) in the range of 400–2,400 nm was investigated in terms of the prediction of heavy metals and mapping of their distribution. Parameters such as absorption depth, area, and peak ratio were derived from variations in spectral absorption shape associated with concentrations of heavy metals. Heavy metals were quantified from SAFPs using stepwise multiple linear regression (SMLR) and enter multiple linear regression (EMLR). The EMLR model showed qualitative prediction performance for As and Cu, with R ² values of 0.60 and 0.81, respectively. A contour map of As and Cu concentrations based on EMLR-derived values showed similar spatial patterns to a map based on measured values.