Prediction of heavy metals in polluted mangrove soils in Brazil with the highest reported levels of mercury using near-infrared spectroscopy
Fábia Shirley Ribeiro Silva,Yuri Jacques Agra Bezerra da Silva,Angelo Jamil Maia,Caroline Miranda Biondi,Paula Renata Muniz Araújo,Ronny Sobreira Barbosa,Cinthia Maria Cordeiro Atanázio Cruz Silva,Tereza Cristina Sidrone Luiz,Arthur Felipe Valença Araújo,Cinthia Maria Cordeiro Atanázio Cru Silva
DOI: https://doi.org/10.1007/s10653-023-01717-2
2023-08-22
Environmental Geochemistry and Health
Abstract:Infrared reflectance spectroscopy has demonstrated potential as a tool for monitoring and preventing contamination in different environments. The objective of this study was to evaluate the usage of near-infrared spectroscopy for predicting heavy-metal contamination in mangrove soils from the Botafogo River estuary located in Pernambuco State, Northeastern Brazil. These soils exhibit the highest mercury (Hg) levels ever reported for Brazilian mangrove soils. Sixty-one samples (obtained at depths ranging from 0 to 5 cm) were collected and measured using near-infrared (1000–2500 nm) reflectance spectroscopy. Preprocessing methods were applied, and partial least squares regression was used to build prediction models for attributes such as clay content, soil organic matter (SOM), pH, Eh, and concentrations of Cr, Cu, Hg, Ni, Pb, and Zn. The models were evaluated using root mean squared error (RMSE), the adjusted coefficient of determination (R 2 adj ), bias, the ratio of performance to interquartile distance (RPIQ), and Lin's concordance correlation coefficient (CCC). The best outcomes were noted for concentrations of Cr, Cu, Hg, Ni, and Pb (RPIQ > 2.5 and R 2 adj > 0.80); second-best outcomes were found for Zn and SOM (RPIQ > 1.5 and R 2 adj > 0.70). Clay content, pH and Eh exhibited the poorest outcomes (RPIQ < 1.5). The importance of spectral preprocessing is highlighted, notably with Savitzky-Golay derivatives and Multiplicative Scatter Corrections, which boosted performance for most of the variables. Near-infrared spectroscopy can be efficiently used to predict Cr, Cu, Hg, Ni, Pb and SOM and represents a technique complementary to traditional analyses.
environmental sciences,engineering, environmental,water resources,public, environmental & occupational health