Three-dimensional quantitative source apportionment of soil heavy metal(loid)s in a mining area

In this study, the three-dimensional (3D) spatial distributions of soil heavy metal(loid)s (HMs) were generated using a hybrid method combining 3D kriging with a depth function accounting for heterogeneity (3DK_DF), and 3D source apportionment was conducted using the principal component analysis with absolute principal component scores. A mining and industrial cluster area in the eastern part of Daye, Hubei Province was selected as study area. Results indicated that As, Cd, Co, Cu, Pb, and Zn were predominantly derived from anthropogenic sources, whereas Cr and Ni primarily originated from natural sources. Furthermore, as anthropogenic sources encompass multiple specific pollution types, a neural network model was developed to further quantify their contributions to soil HMs. Results revealed that mines and tailings ponds were the primary contributors to soil As, Cd, Co, Cu, Pb, and Zn within the 0-80 cm depth range, with contribution percentages of 26.50 %-31.78 % and 23.48 %-30.38 %, and total contribution percentages of 26.91 % and 26.06 %, respectively. Moreover, the migration and diffusion of HMs in 3D soil generally exhibited outward and downward trends from pollution sources. However, their concentration patterns of variation exhibited notable differences depending on the source type and HM species. The proposed integration of 3D spatial interpolation and a specific anthropogenic source apportionment model can be used to distinguish the major anthropogenic sources responsible for soil HMs pollution, providing data support for the precise identification and effective control of pollution types in mining areas.