Using geostatistical modeling methods to assess concentration and spatial variability of trace metals in soils of the abandoned gold mining district of Bindiba (East Cameroon)
Martin Mozer Njayou,Mouhamed Ngounouno Ayiwouo,Luc Leroy Ngueyep Mambou,Ismaïla Ngounouno
DOI: https://doi.org/10.1007/s40808-022-01560-x
2022-10-20
Modeling Earth Systems and Environment
Abstract:Trace metal pollution in surface soils is of special concern given the potential dangers to human health. This paper uses geostatistical modelling to assess the spatial variability of soil physico-chemical parameters (pH, EC) and trace metals (Cr, Ni, Cu, As and Pb) in the abandoned gold mining district of Bindiba (East Cameroon). Two sampling campaigns are carried out in the dry season and rainy season and a total of 89 samples are collected at an average spacing of 50 m. To produce realistic prediction maps at un-sampled locations, geostatistical analysis is used. The geostatistical approach use exploratory analysis, variographic analysis (VA), and ordinary kriging (OK). The soils of the abandoned gold mining district are characterized by acidic to near neutral pH (5.01–6.19), weakly conductivities (7–47 µS cm−1) and trace metals range from Pb (0.006–53.27 mg kg−1), As (0.00–46.58 mg kg−1), Cr (22.15–442.44 mg kg−1), Ni (9.25–360.37 mg kg−1) and Cu (1.28–320.86 mg kg−1). The variographic analysis of pH, EC and trace metals which highlight the heterogeneity of the contamination, reveal two variogram models, spherical model for Cr, As, Pb, pH and EC, and exponential model for Ni and Cu. The trace metals Pb (1.95) and As (1.78) show the highest variability, while the lowest variability is observed for Cu (0.37). The length of the spatial autocorrelation is much longer than the sampling step indicationg that the sampling design adopted in this study is appropriate. The nugget/sill ratio values of 0.65, 0.67, 0.62, 0.27, 0.57 and 0.19 for pH, EC, Cr, Ni, As and Pb, respectively, suggest a moderate spatial dependence. High concentrations of potentially toxic elements are found in the soils of the study area, indicating that anthropogenic factors are causing the anomalies in these areas. The maps obtained from the geostatistical modelling accurately described the spatial variability of trace metal concentrations in soil. Thus, this study could help decision makers to develop a better soil management strategy.