Search

Detailed mapping of soil attributes is often not viable due to the high cost of wet‐chemical laboratory analysis, which requires a large number of samples. Thus, we evaluated whether the prediction of SOC contents through field‐specific diffuse reflectance spectroscopy (DRS) can increase the amount of samples available to SOC mapping through data interpolation. For such, we tested the performance of the partial least squares regression (PLSR), random forest (RF) and gradient boosting tree (GBT) algorithms to model and predict SOC. The field‐specific calibration approach proposed here proved to be suitable for predicting SOC content on soil samples, reducing the dependence on wet‐chemical soil laboratory analyses for mapping. With such SOC content prediction, the higher amount of samples to be used for spatial interpolation can be increased, leading to more accurate SOC maps that can be applied for site‐specific management. [ABSTRACT FROM AUTHOR]

Environmental impacts can be more accurately assessed with the aid of spatial characterization of potentially toxic elements (PTEs). In fact, developing cost-effective, environmentally friendly spatial characterization methods for PTEs can facilitate the expeditious, accurate, detailed diagnosis of soil in large areas. In this study, we used three geomorphic surfaces of Oxisols to assess the ability of chemical and X-ray diffraction analyses of iron oxides, diffuse reflectance spectroscopy (DRS) and magnetic susceptibility (MS) measurements to predict the contents in PTEs (Ba, Co, Cr, Cu, Mn, Mo, Ni, Pb and Zn) and their spatial variability. Magnetic susceptibility and diffuse reflectance spectroscopy allowed well-calibrated prediction models for Ba, Co, Cu, Mn and Ni to be developed, whereas DRS-calibrated methods afforded more accurate prediction of Ba and Mn contents, and magnetic susceptibility-calibrated methods of Co and Ni contents. The correlation between PTEs and free iron contents, and their spatial pattern, testifies to the goodness of the proposed methods for predicting the contents of potentially toxic elements in soils. [ABSTRACT FROM AUTHOR]

The development of fast, accurate and low-cost methods to quantify soil attributes is of paramount importance to enable detailed mapping, mainly in tropical regions where there is great variation of the chemical, physical and mineralogical attributes. Therefore, the aims of this paper were (i) to investigate if visible and near infrared (VIS-NIR) spectroscopy and magnetic susceptibility (MS) can be applied to determine soil attributes at the sandstone-basaltic transition and (ii) evaluate and map their spatial distribution. Calibration models based on VIS-NIR spectroscopy and MS were developed separately for each attribute. Soil samples (0-25 cm depth) were collected at 446 sites, air-dried and passed through a 2-mm sieve and analyzed in the laboratory. To develop models based on soil spectra and laboratory data, the partial least squares regression (PLSR) was used. Already, the MS-based models were calibrated by linear regression between magnetic and laboratory data. The best prediction accuracy parameters were obtained with MS, later with VIS-NIR and lastly with VIS. The more accurate results between the observed and predicted values were found for iron oxide extracted by dithionite (R² = 0.89, RRMSE = 0.02), clay (R² = 0.85, RRMSE = 0.76) and total carbon (R² = 0.83, RRMSE = 1.18) estimated by MS, revealing that this is a good predictor of key properties of studied soils, even with wide chemical and mineralogical variation. Both tools are very attractive for the strategic planning of land use and occupation, mapping large areas with detailed scale, environmental monitoring and precision agriculture. [ABSTRACT FROM AUTHOR]

Diffuse reflectance spectroscopy (DRS) provides a rapid, non-destructive, less expensive and sustainable complement to conventional methods of soil analysis. This study aims to assess the DRS ability to predict contents of clay, iron oxide and adsorbed phosphate (Pads), characterizing their spatial variability in Oxisols from Brazil. Soil samples collected from three geomorphic surfaces were scanned in the visible and near infrared spectral range. Moreover, chemometric calibration was obtained through partial least-squares regression (PLSR). Then, spatial distribution maps of the attributes were constructed using predicted values from calibrated models through geostatistical methods. The best accuracy of prediction models (residual prediction deviation > 2.0) for clay, extractable Fe by dithionite–citrate–bicarbonate and hematite was obtained within the visible region (380–800 nm), and for adsorbed phosphate within the visible near infrared (380–2300 nm). Predicted values of these attributes had low error (< 10%) with the exception of goethite, and enabled building maps similar to those with measured values, demonstrating the DRS importance for clay, iron oxide and adsorbed phosphate mapping of large areas and at detailed scale. [ABSTRACT FROM AUTHOR]