Accurately mapping the spatial distribution information of soil organic carbon (SOC) stocks is a key premise for soil resource management and environment protection. Rapid development of satellite remote sensing provides a great opportunity for monitoring SOC stocks at a large scale. In this study, based on 12 environmental variables of multispectral remote sensing, topography and climate and 236 soil sampling data, three different boosted regression tree (BRT) models were compared to obtain the most accurate map of SOC stocks covering the forest area of Lvshun District in the Northeast China. Four validation indexes, including mean absolute error (MAE), root mean square error (RMSE), coefficient of determination (R²), and Lin’s concordance correlation coefficient (LCCC) were calculated to evaluate the performance of the three models. The results showed that the full variable model performed the best, except the model using multispectral remote sensing variables. In the full variable model, the regional SOC stocks are primarily determined by multispectral remote sensing variables, followed by topographic and climatic variables, with the relative importance of variables in the model being 63%, 28%, and 9%, respectively. The average prediction results of full variables model and only multispectral remote sensing variables model were 8.99 and 9.32 kg m⁻², respectively. Our results indicated that there is a strong dependence of SOC stocks on multispectral remote sensing data when forest ecosystems have dense natural vegetation. Our study suggests that the multispectral remote sensing variables should be used to map SOC stocks of forest ecosystems in our study region.

Soil organic matter (SOM) is a key index of soil fertility. Visible and near-infrared (VNIR, 350–2500 nm) reflectance spectroscopy is an effective method for modeling SOM content. Characteristic wavelength screening and spectral transformation may improve the performance of SOM prediction. This study aimed to explore the optimal combination of characteristic wavelength selection and spectral transformation for hyperspectral modeling of SOM. A total of 219 topsoil (0–20 cm) samples were collected from two soil types in the East China. VNIR reflectance spectra were measured in the laboratory. Firstly, after spectral transformation (inverse-log reflectance (LR), continuum removal (CR) and first-order derivative reflectance (FDR)) of VNIR spectra, characteristic wavelengths were selected by competitive adaptive reweighted sampling (CARS) and uninformative variables elimination (UVE) algorithms. Secondly, the SOM prediction models were constructed based on the partial least squares regression (PLSR), random forest (RF) and support vector regression (SVR) methods using the full spectra and selected wavelengths, respectively. Finally, optimal SOM prediction models were selected for two soil types. The results were as follows: (1) The CARS algorithm screened 40–125 characteristic wavelengths from the full spectra. The UVE algorithm screened 105–884 characteristic wavelengths. (2) For two soil types and full spectra, CARS and UVE improved the SOM modeling precision based on the PLSR and SVR methods. The coefficient of determination (R²) value in the validation of the CARS-PLSR (PLSR model combined with CARS) and CARS-SVR (SVR model combined CARS) models ranged from 0.69 to 0.95, and the relative percent deviation (RPD) value ranged from 1.74 to 4.31. Lin’s concordance correlation coefficient (LCCC) values ranged from 0.83 to 0.97. The UVE-PLSR and UVE-SVR models showed moderate precision. (3) The PLSR and SVR modeling accuracies of Paddy soil were better than those for Shajiang black soil. RF models performed worse for both soil types, with the R² values of validation ranging from 0.22 to 0.68 and RPD values ranging from 1.01 to 1.60. (4) For Paddy soil, the optimal SOM prediction models (highest R² and RPD, lowest root mean square error (RMSE)) were CR-CARS-PLSR (R² and RMSE: 0.97 and 1.21 g/kg in calibration sets, 0.95 and 1.72 g/kg in validation sets, RPD: 4.31) and CR-CARS-SVR (R² and RMSE: 0.98 and 1.04 g/kg in calibration sets, 0.91 and 2.24 g/kg in validation sets, RPD: 3.37). For Shajiang black soil, the optimal SOM prediction models were LR-CARS-PLSR (R² and RMSE: 0.95 and 0.93 g/kg in calibration sets, 0.86 and 1.44 g/kg in validation sets, RPD: 2.62) and FDR-CARS-SVR (R² and RMSE: 0.99 and 0.45 g/kg in calibration sets, 0.83 and 1.58 g/kg in validation sets, RPD: 2.38). The results suggested that the CARS algorithm combined CR and FDR can significantly improve the modeling accuracy of SOM content.

This study aimed to evaluate and compare the performances of the random forest (RF) and support vector regression (SVR) models combined with different feature selection methods, including recursive feature elimination (RFE), simulated annealing feature selection (SAFS), and selection by filtering (SBF) in predicting soil pH in Anhui Province, East China. We also used the ALL original features to build the RF and SVR models as a comparison. A total of 140 samples were selected, following the principles of randomness, uniformity, and representativeness, to consider the combination of landscape elements, such as topography, parent material, and land use. Auxiliary data, including climatic, topographic, and vegetation indexes, were used for predicting soil pH. The results showed that compared with the use the ALL original modeling features (ALL-RF, ALL-SVR), the combination of the three feature selection algorithms with RF and SVR can eliminate some redundant features and effectively improve the prediction accuracy of the soil pH model. For the RF model, the RMSE and the MAE of the calibration of the RFE-RF model were 0.73 and 0.57 and had the highest R² in four different RF models. The testing set of the RFE-RF model had an R² of 0.61, which was better than that of the ALL-RF (R² = 0.45) model and lower than those of the SAFS-RF (R² = 0.71) and SBF-RF (R² = 0.69) models. For the SVR model, the RFE-RF model was more robust and had better generalization ability. The accuracy of digital soil mapping can be improved through feature selection.

The mineral composition of Fe–Mn nodules (FMNs) is closely related to the environment in which FMNs are formed. The formation mechanism of FMNs has consistently been one of the major topics in related research. In this study, the mineral composition of FMNs in soils derived from carbonate rocks in typical karst areas with high geochemical background in Guangxi, Southwest China, was investigated. The results showed that Fe₂O₃ (30.06%), SiO₂ (19.72%), Al₂O₃ (17.93%), TiO₂ (0.96%), P₂O₅ (0.78%), and MnO₂ (0.64%) were the main elemental composition, while four alkaline oxides K₂O, Na₂O, CaO, and MgO were less than 0.5% in soil FMNs. In addition, 10 mineral types were identified in soil FMNs in the study area, namely quartz, goethite, clinochlore, illite, kaolinite, boehmite, albite, microcline, lithiophorite, and hematite. There were no obvious differences in the mineral composition of soil FMNs in the study area compared with those observed in non-karst areas worldwide. The formation process of FMNs can be determined based on the surrounding environment that affects the soil mineral composition. The results suggested a relatively complex formation mechanism of soil FMNs. Moreover, both primary and secondary minerals were found in soil FMNs. However, some minerals can exist stably under normal redox conditions, while the other part of minerals can be easily weathered and dissolved, indicating a relatively high formation rate of soil FMNs and a relatively stable internal environment.

Nanoscale zero-valent iron (nZVI) has been broadly applied in the remediation of heavy metals pollution. In this research, the toxicity characteristic leaching procedure (TCLP), the in vitro gastrointestinal (IVG) method, and the diffusive gradients in thin-films (DGT) technique were used to evaluate the effects of heavy metals in soil with remediation by nZVI. The results indicate that, compared with the dose of 0.5 g·L⁻¹, the nZVI in the dose of 1.0 g·L⁻¹ can remove the heavy metals in the soil. The leaching toxicities of the heavy metals (Cr, Cu, Zn, Pb) showed apparent decreases after the remediation by nZVI. In the gastric phase, the highest bioaccessibility values of the Cr, Cu, Zn, Pb were decreased by 27.2, 31.7, 11.7, and 20.1%, respectively. Moreover, in the gastric phase, the highest bioaccessibility values of the Cr, Cu, Zn, Pb were decreased by 5.5, 1.29, 8.0, and 7.3%, respectively. The availabilities of the heavy metals were also reduced. The above results show that the nZVI effectively reduced the heavy metal pollution in the soil.

Heavy metal pollution in agricultural soil has been tightly associated with anthropogenic emissions. Although there are many studies that focus on a regional scale, the source identification of heavy metal contamination on a field scale around industrial areas remains unclear. The average concentrations in topsoils of Hg, Cd, As, Pb, Cr, Ni, Zn, and Cu were 2.07, 0.13, 8.56, 42.3, 81.1, 37.3, 105, and 43.8 mg kg⁻¹, respectively. The enrichment of Hg was particularly presented on topsoils, with the highest single pollution index (Pi) (9.00) and ecological risk index (Eri) (922) values. An integrated methodology was employed in source identification of heavy metals contamination, especially for Hg, including Pearson’s and PCA analysis, soil profile morphology, mathematical modeling, and Hg isotope analysis. Results revealed that the concentrations of Hg decreased as a function of depth, suggesting Hg contamination was an anthropogenic source and can be supported by Hg isotope analysis. The negative Δ¹⁹⁹Hg values of the residual Hg (F4-Hg) and soil profile in 80–100 cm deviate from those of the soil profiles in 0–80 cm, indicating exogenous input of Hg occurred in the study area. According to the UNMIX model, the contribution of coal combustion, agricultural activities, parent material, and industrial/traffic emissions to Hg accumulation in soils were 66.2%, 16.9%, 9.81%, and 7.0%, respectively. However, the contribution rates calculated with the PMF model of mixed industrial source, traffic emissions, and parent material were 71.4%, 27.8%, and 0.8%, respectively. This study can accurately quantify and identify the factors contributing to heavy metal contamination in agricultural soil on a field scale.