Estimation and Spatial Analysis of Heavy Metals in Metal Tailing Pond Based on Improved PLS With Multiple Factors
Wang Rui,Wu Shuang,Wu Kan,Huang Shiqiao,Wu Ruijie,Liu Bo,Lin Min,Li Liang,Zhou Dawei,Diao Xinpeng
DOI: https://doi.org/10.1109/access.2021.3073933
IF: 3.9
2021-01-01
IEEE Access
Abstract:With the exploitation of mineral resources, pollution of the ecological environment in mines has garnered public attention. Particularly,erosion of the surrounding ecological environment re-sulting from heavy metals in tailings pond could be highly concerning. Instead of traditional field sampling and laboratory analysis method, remote sensing can be used to high-precisi es-timation soil heavy metal with less time and effort. soil heavy metal content is generally low, the spectral sensitivities of various heavy metals are insignificant, and the surface landscape is complex, there exist difficulties associated with heavy metal content estimation. Therefore, herein, we propose optimization of the commonly used partial least-square regression (PLS) method. In the optimized method, a variety of remote sensing indices and the modeled heavy metals were added as modeling factors to indirect estimation soil heavy metal. The method was validated via inversion experiments of heavy metals (Ni, Cu, and Zn) in the tailing pond and its surrounding environment,it improve the goodness-of-fit of Ni, Cu, and Zn by 0.0852,0.2291, and 0.2919 compared with traditional PLS. Spatia l analysis was then conducted on the entire studied area using the estimation model of the three heavy metals. It was shown that the results were essentially consistent with the actual heavy metal distribution in the area. Therefore, the indirect PLS model with multiple factors proves effective for the estimation of soil heavy metals. It also provides technical support for treatment and evaluation of ecological environments in mining areas.
computer science, information systems,telecommunications,engineering, electrical & electronic