Search

Leaf optical properties can be used to identify environmental conditions, the effect of light intensities, plant hormone levels, pigment concentrations, and cellular structures. However, the reflectance factors can affect the accuracy of predictions for chlorophyll and carotenoid concentrations. In this study, we tested the hypothesis that technology using two hyperspectral sensors for both reflectance and absorbance data would result in more accurate predictions of absorbance spectra. Our findings indicated that the green/yellow regions (500–600 nm) had a greater impact on photosynthetic pigment predictions, while the blue (440–485 nm) and red (626–700 nm) regions had a minor impact. Strong correlations were found between absorbance (R2 = 0.87 and 0.91) and reflectance (R2 = 0.80 and 0.78) for chlorophyll and carotenoids, respectively. Carotenoids showed particularly high and significant correlation coefficients using the partial least squares regression (PLSR) method (R2C = 0.91, R2cv = 0.85, and R2P = 0.90) when associated with hyperspectral absorbance data. Our hypothesis was supported, and these results demonstrate the effectiveness of using two hyperspectral sensors for optical leaf profile analysis and predicting the concentration of photosynthetic pigments using multivariate statistical methods. This method for two sensors is more efficient and shows better results compared to traditional single sensor techniques for measuring chloroplast changes and pigment phenotyping in plants.

Unstable soil aggregates can easily be dispersed when mechanical energy is applied, denominated as readily dispersible clay (RDC) which is usually measured by turbidimetry. Faster methods that requ...

International audience; Spectroscopic measurements of soil samples are reliable because they are highly repeatable and reproducible. They characterise the samples' mineral-organic composition. Estimates of concentrations of soil constituents are inevitably less precise than estimates obtained conventionally by chemical analysis. But the cost of each spectroscopic estimate is at most one-tenth of the cost of a chemical determination. Spectroscopy is cost-effective when we need many data, despite the costs and errors of calibration. Soil spectroscopists understand the risks of over-fitting models to highly dimensional multivariate spectra and have command of the mathematical and statistical methods to avoid them. Machine learning has fast become an algorithmic alternative to statistical analysis for estimating concentrations of soil constituents from reflectance spectra. As with any modelling, we need judicious implementation of machine learning as it also carries the risk of over-fitting predictions to irrelevant elements of the spectra. To use the methods confidently, we need to validate the outcomes with appropriately sampled, independent data sets. Not all machine learning should be considered 'black boxes'. Their interpretability depends on the algorithm, and some are highly interpretable and explainable. Some are difficult to interpret because of complex transformations or their huge and complicated network of parameters. But there is rapidly advancing research on explainable machine learning, and these methods are finding applications in soil science and spectroscopy. In many parts of the world, soil and environmental scientists recognise the merits of soil spectroscopy. They are building spectral libraries on which they can draw to localise the modelling and derive soil information for new projects within their domains. We hope our article gives readers a more balanced and optimistic perspective of soil spectroscopy and its future. Highlights Spectroscopy is reliable because it is a highly repeatable and reproducible analytical technique. Spectra are calibrated to estimate concentrations of soil properties with known error. Spectroscopy is cost-effective for estimating soil properties. Machine learning is becoming ever more powerful for extracting accurate information from spectra, and methods for interpreting the models exist. Large libraries of soil spectra provide information that can be used locally to aid estimates from new samples.

Retaining a residue cover (RC) over the soil is the key to ensure its health. Since tropical agriculture systems, mostly in sugarcane, are characterized by intensive management, the discrimination ...

Applying the upcoming technologies in agriculture has been a major economic, environmental and social challenge for scientists and farmers. In order to overcome such challenge, this study evaluated the advantages and limitations of using geostatistics and machine learning for soil mapping in agricultural practices and soil surveys. The study occurred in Tocantins State, Brazil, and consisted into seven areas with a total extension of 17.24 km2, 222 meters regular gridded resulting in one-point sampling per 0.0493 km2 of five randomly sampled cores within a 1 m circle radius. It was collected 332 georeferenced soil samples at 0-20 cm depth using an auger and then, soil laboratory analyses performed. Afterward, liming rate maps were originated from the predicted soil attributes clay, cation exchange capacity and base saturation comparing four methods: ordinary kriging, random forest, cubist, support vector machine and the best model results of each soil attribute. Evaluating the methods, the Pearson’s index presented strong results for soil attributes predicted by random forest and ordinary kriging. Machine learning methods can be successfully applied for soil mapping in agricultural practices and soil surveys using less soil samples rather than geostatistical framework.

The present study was conducted to evaluate the effectiveness of combining proximal, and remote sensing with environmental variables for predicting USDA (United States Department of Agriculture) so...

Soil classification has traditionally been developed by combining the interpretation of taxonomic rules that are related to soil information with the pedologist's tacit knowledge. Hence, a more quantitative approach is necessary to characterize soils with less subjectivity. The objective of this study was to develop a soil grouping system based on spectral, climate, and terrain variables with the aim of establishing a quantitative way of classifying soils. Spectral data were utilized to obtain information about the soil, and this information was complemented by climate and terrain variables in order to simulate the pedologist knowledge of soil–environment interactions. We used a data set of 2287 soil profiles from five Brazilian regions. The soil classes of World Reference Base (WRB) system were predicted using the three above-mentioned variables, and the results showed that they were able to correctly classify the soils with an overall accuracy of 88 %. To derive the new system, we applied the spectral, climatic, and terrain variables, which – using cluster analysis – defined eight groups; thus, these groups were not generated by the traditional taxonomic method but instead by grouping areas with similar characteristics expressed by the variables indicated. They were denominated as “soil environment groupings” (SEGs). The SEG system facilitated the identification of groups with equivalent characteristics using not only soil but also environmental variables for their distinction. Finally, the conceptual characteristics of the eight SEGs were described. The new system has been designed to incorporate applicable soil data for agricultural management, to require less interference from personal/subjective/empirical knowledge (which is an issue in traditional taxonomic systems), and to provide more reliable automated measurements using sensors.

Detailed soil maps contribute to the understanding of agricultural and food value chains. PlanetScope satellite constellations have high spatial and temporal resolution and could be used to develop...

Geophysical sensors combined with machine learning algorithms were used to understand the pedosphere system and landscape processes and to model soil attributes. In this research, we used parent material, terrain attributes, and data from geophysical sensors in different combinations to test and compare different and novel machine learning algorithms to model soil attributes. We also analyzed the importance of pedoenvironmental variables in predictive models. For that, we collected soil physicochemical and geophysical data (gamma-ray emission from uranium, thorium, and potassium; magnetic susceptibility and apparent electric conductivity) by three sensors (gamma-ray spectrometer, RS 230; susceptibilimeter KT10, Terraplus; and conductivimeter, EM38 Geonics) at 75 points and analyzed the data. The models with the best performance (R2 0.48, 0.36, 0.44, 0.36, 0.25, and 0.31) varied for clay, sand, Fe2O3, TiO2, SiO2, and cation exchange capacity prediction, respectively. Modeling with the selection of covariates at three phases (variance close to zero, removal by correction, and removal by importance) was adequate to increase the parsimony. The results were validated using the method “nested leave-one-out cross-validation”. The prediction of soil attributes by machine learning algorithms yielded adequate values for field-collected data, without any sample preparation, for most of the tested predictors (R2 values ranging from 0.20 to 0.50). Also, the use of four regression algorithms proved to be important since at least one of the predictors used one of the tested algorithms. The performance values of the best algorithms for each predictor were higher than those obtained with the use of a mean value for the entire area comparing the values of root mean square error (RMSE) and mean absolute error (MAE). The best combination of sensors that reached the highest model performance was that of the gamma-ray spectrometer and the susceptibilimeter. The most important variables for most predictions were parent material, digital elevation, standardized height, and magnetic susceptibility. We concluded that soil attributes can be efficiently modeled by geophysical data using machine learning techniques and geophysical sensor combinations. This approach can facilitate future soil mapping in a more time-efficient and environmentally friendly manner.

Soil salinization is a severe danger to agricultural activity in arid and semi-arid areas, reducing crop production and contributing to land destruction. This investigation aimed to utilize machine learning algorithms to predict spatial soil salinity (dS m−1) by combining environmental covariates derived from remotely sensed (RS) data, a digital elevation model (DEM), and proximal sensing (PS). The study is located in an arid region, southern Iran (52°51′–53°02′E; 28°16′–28°29′N), in which we collected 300 surface soil samples and acquired the spectral data with RS (Sentinel-2) and PS (electromagnetic induction instrument (EMI) and portable X-ray fluorescence (pXRF)). Afterward, we analyzed the data using five machine learning methods as follows: random forest—RF, k-nearest neighbors—kNN, support vector machines—SVM, partial least squares regression—PLSR, artificial neural networks—ANN, and the ensemble of individual models. To estimate the electrical conductivity of the saturated paste extract (ECe), we built three scenarios, including Scenario (1): Synthetic Soil Image (SySI) bands and salinity indices derived from it; Scenario (2): RS data, PS data, topographic attributes, and geology and geomorphology maps; and Scenario (3): the combination of Scenarios (1) and (2). The best prediction accuracy was obtained for the RF model in Scenario (3) (R2 = 0.48 and RMSE = 2.49), followed by Scenario (2) (RF model, R2 = 0.47 and RMSE = 2.50) and Scenario (1) for the SVM model (R2 = 0.26 and RMSE = 2.97). According to ensemble modeling, a combined strategy with the five models exceeded the performance of all the single ones and predicted soil salinity in all scenarios. The results revealed that the ensemble modeling method had higher reliability and more accurate predictive soil salinity than the individual approach. Relative improvement (RI%) showed that the R2 index in the ensemble model improved compared to the most precise prediction for the Scenarios (1), (2), and (3) with 120.95%, 56.82%, and 66.71%, respectively. We applied the best model in each scenario for mapping the soil salinity in the selected area, which indicated that ECe tended to increase from the northwestern to south and southeastern regions. The area with high ECe was located in the regions that mainly had low elevations and playa. The areas with low ECe were located in the higher elevations with steeper slopes and alluvial fans, and thus, relief had great importance. This study provides a precise, cost-effective, and scientific base prediction for decision-making purposes to map soil salinity in arid regions.

We conducted a systematic review and inventory of recent research achievements related to spaceborne and aerial Earth Observation (EO) data-driven monitoring in support of soil-related strategic goals for a three-year period (2019–2021). Scaling, resolution, data characteristics, and modelling approaches were summarized, after reviewing 46 peer-reviewed articles in international journals. Inherent limitations associated with an EO-based soil mapping approach that hinder its wider adoption were recognized and divided into four categories: (i) area covered and data to be shared; (ii) thresholds for bare soil detection; (iii) soil surface conditions; and (iv) infrastructure capabilities. Accordingly, we tried to redefine the meaning of what is expected in the next years for EO data-driven topsoil monitoring by performing a thorough analysis driven by the upcoming technological waves. The review concludes that the best practices for the advancement of an EO data-driven soil mapping include: (i) a further leverage of recent artificial intelligence techniques to achieve the desired representativeness and reliability; (ii) a continued effort to share harmonized labelled datasets; (iii) data fusion with in situ sensing systems; (iv) a continued effort to overcome the current limitations in terms of sensor resolution and processing limitations of this wealth of EO data; and (v) political and administrative issues (e.g., funding, sustainability). This paper may help to pave the way for further interdisciplinary research and multi-actor coordination activities and to generate EO-based benefits for policy and economy.

Arsenic (As) and lead (Pb) are potentially toxic elements capable of developing several diseases in human beings such as cancer. Several adsorbent materials, including biochars, have been adopted as alternative measures designed to reduce the availability of As and Pb in water. The retention capacity of potentially toxic elements in biochars varies according to time, feedstock, and the pyrolysis temperature to produce the biochar. Our objectives in this study were to evaluate i) the aging effect of sugarcane straw pyrolyzed biochars at 350 (BC350), 550 (BC550), and 750 °C (BC750) and their ability to immobilize As and Pb; and ii) how the pyrolysis temperature and biochar aging alter the carbon content and quality of the solution and sediment. Biochars were applied at 5% (w/w), and their aging together with As and Pb immobilization effects were evaluated every 45 days over a total period of 180 days. The results were obtained using visible ultraviolet spectroscopy and diffuse reflectance infrared Fourier transform spectroscopy combined with physical fractionation of organic matter and multivariate statistics. The groups formed in the Principal Component Analysis indicated that the change in the availability of As and Pb was related to the aging of the biochar and the temporal changes in the content and quality of organic carbon in the sediment and solution. The pyrolysis temperature was a key factor in the (im)mobilization capacity of As and Pb during the aging of the biochar. The increase in polysaccharides and organic matter associated with the particulate fraction can enhance the release of As in solution (24%). Increasing the fraction of organic matter associated with minerals reduced the availability of Pb by 58%. These findings may provide new insights into understanding the dynamics of organic matter and its role in the immobilization of As and Pb during biochar aging.

Geophysical sensors combined with machine learning algorithms have been used to understand the pedosphere system, landscape processes and to model soil attributes. In this research, we used parent material, terrain attributes and data from geophysical sensors in different combinations, to test and compare different and novel machine learning algorithms to model soil attributes. Also, we analyzed the importance of pedoenvironmental variables in predictive models. For that, we collected soil physico-chemical and geophysical data (gamma-ray emission from uranium, thorium and potassium, magnetic susceptibility and apparent electric conductivity) by three sensors, gamma-ray spectrometer – RS 230, susceptibilimeter KT10 – Terraplus and Conductivimeter – EM38 Geonics) at 75 points and, we performed soil analysis afterwards. The results showed varying models with the best performance (R2 > 0.2) for clay, sand, Fe2O3, TiO2, SiO2 and Cation Exchange Capacity prediction. Modeling with selection of covariates at three phases (variance close to zero, removal by correction and removal by importance), demonstrated to be adequate to increase the parsimony. The prediction of soil attributes by machine learning algorithms demonstrated adequate values for field collected data, without any sample preparation, for most of the tested predictors (R2 ranging from 0.20 to 0.50). Also, the use of four regression algorithms proved important, since at least one of the predictors used one of the tested algorithms. The performances of the best algorithms for each predictor were higher than the use of a mean value for the entire area comparing the values of Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The best combination of sensors that reached the best model performance to predict soil attributes were gamma-ray spectrometer and susceptibilimeter. The most important variables were parent material, digital elevation model, standardized height and magnetic susceptibility for most predictions. We concluded that soil attributes can be efficiently modelled by geophysical data using machine learning techniques and geophysical sensors combinations. The technique can bring light for future soil mapping with gain of time and environment friendly.

It is impossible to make pedological maps without understanding subsurface attributes. Several strategies can be used for soil mapping, from a tacit knowledge to mathematical modeling. However, there are still gaps in knowledge regarding how to optimize subsurface mapping. This work aimed to quantify subsurface soil attributes using satellite spectral reflectance and geographically weighted regression (GWR) techniques. The study was carried out in Sao Paulo, Brazil, in an area spanning 47,882 ha. Multitemporal satellite images (Landsat-5) were initially processed in order to retrieve spectral reflectance from the bare soil surface. Based on a toposequence method, 328 points were then distributed across the area (at depths between 0 and 20 cm and 80 and 100 cm) and analyzed for their soil chemical and physical attributes (including the reflectance spectra (400 to 2500 nm)) in the laboratory. We achieved 67.72% of bare soil for the whole study area, with the remaining 32.28% of the unmapped surface being filled by kriging interpolation. All 328 samples were modeled using surface (Landsat-5 TM spectral reflectance) and subsurface (acquired in the laboratory) data, reaching up to 0.72 R2adj. The correlation between the spectra of both depths was significant and the soil attributes prediction reached an R2adj of validation above 0.6 for clay, hue, value, and chroma at 0–20 and 80–100 cm depths. The satellite soil surface reflectance allowed the estimation of soil subsurface attributes. These results demonstrate that diagnostic soil attributes can be quantified based on spectral pedotransfer (SPEDO) functions to assist digital soil mapping and soil monitoring. Despite our efforts to determine soil subsurface properties using digital soil mapping approach, this task still need considerable refinement. Thus, research must continue to aggregate outcomes from other techniques.

Considering the hypothesis that the predictive capacity of models is tied to soil characteristics, the stratification of a spectral library into groups is a strategy to improve the accuracy of the predictions. Thus, the objective of this study was to i) characterize and identify differences among spectra obtained for subtropical soils samples, ii) evaluate different pre-processing techniques and multivariate methods to propose SOC prediction models from the spectral data and iii) evaluate the performance of SOC prediction models calibrated from the stratification of a local library. A local spectral library of soils (n = 841 samples) was used in the Planalto region of the State of Rio Grande do Sul, Brazil. Soil classes that occur in the area are: Rhodic Ferralsol (FR) and Dystric Gleysol (GL). Land uses are: native forest (NFo), native field (NFi) and crops in no-tillage system (CTS). SOC was determined via wet combustion with sulphochromic solution. Spectral reflectance measurements were performed in the laboratory with a spectroradiometer in the range of 350–2500 nm. Six pre-processing techniques were applied to the spectra (including derivatives, normalization and non-linear transformations) and four multivariate calibration methods, namely, partial least squares regression (PLSR), multiple linear regression (MLR), support vector machines (SVM) and random forest (RF), were used with the objective of identifying the best combination to predict SOC. After determining the best combination, the spectral library was stratified into groups based on soil class, land use, sample layer and spectral characteristics. The models were built with 70% of the samples for calibration and 30% for independent validation. The coefficient of determination (R2v), root mean square error (RMSEv) and ratio of performance to interquartile range (RPIQv) of the independent validation were used to evaluate the performance of the models. The spectral curves presented different absorption characteristics in relation to soil classes and land uses. SGD pre-processing technique had the highest R2v and RMSEv values for all models. Among the multivariate methods, PLSR had the best performance for SOC prediction for the total set of samples (R2v = 0.74, RMSEv = 0.52% and RPIQv = 2.23), followed by models SVM, MLR, and RF. The FR-CTS (n = 445) group showed the best model performance after stratification, with R2v = 0.82, RMSEv = 0.29% and RPIQv = 2.60. For some stratified groups, the use of a smaller number of samples to build the model reduced the performance of the models, suggesting that one must be careful when working with small datasets. This study highlights the potential for the application of VIS-NIR-SWIR spectroscopy as a reliable and economical tool to quantify SOC concentrations for subtropical soils with high levels of iron oxides and clay on a local scale. Predictive models can be improved when the variation in soil characteristics is considered, underscoring the need for a preliminary study examining the grouping of the sample set to validate the use of local spectral libraries for the prediction of soil properties.

Soil organic carbon (C) is an important indicator of agricultural and environmental quality. It improves soil fertility and helps to mitigate greenhouse gas emissions. Soil spectroscopy with either vis–NIR (350–2500 nm) or mid-IR (4000–400 cm−1) spectra have been used successfully to predict organic C concentrations in soil. However, research to improve predictions of soil organic C by simply combining vis–NIR and mid-IR spectra to model them together has been unsuccessful. Here we use the Outer Product Analysis (OPA) to fuse vis–NIR and mid-IR spectra by bringing them into a common spectral domain. Using the fused data, we derived models to predict soil organic C and compared its predictions to those derived with vis–NIR and mid-IR models separately. We analyzed 1259 tropical soil samples from surface and subsurface layers across agricultural areas in Central Brazil. Soil organic C contents were determined by a modified Walkley-Black method, and vis–NIR and mid-IR reflectance spectra were obtained with a FieldSpec Pro and a Nicolet 6700 Fourier Transformed Infrared (FT-IR), respectively. Reflectances were log-transformed into absorbances. The mean content of soil organic C was 9.14 g kg−1 (SD = 5.64 g kg−1). The OPA algorithm was used to emphasize co-evolutions of each spectral domain into the same one by multiplying the absorbances from both sets of spectra to produce a matrix with all possible products between them. Support Vector Machine with linear kernel function was used for the spectroscopic modeling. Predictions of soil organic C using vis–NIR, mid-IR, and fused spectra were statistically compared by the Tukey's test using the coefficient of determination (R2), root mean squared error (RMSE), and ratio of performance to interquartile distance (RPIQ). Absorbances in vis–NIR and mid-IR were emphasized in the common spectral domain presenting stronger correlations with soil organic C than individual ranges. Soil organic C predictions with the OPA fused spectra were significantly better (R2 = 0.81, RMSE = 2.42 g kg−1, and RPIQ = 2.87) than those with vis–NIR (R2 = 0.69, RMSE = 3.38 g kg−1, and RPIQ = 2.08) or mid-IR spectra (R2 = 0.77, RMSE = 2.90 g kg−1, and RPIQ = 2.43). Fusing vis–NIR and mid-IR spectra by OPA improves predictions of soil organic C.

Visible (V), Near Infrared (NIR) and Short Waves Infrared (SWIR) spectroscopy has been indicated as a promising tool in soil studies, especially in the last decade. However, in order to apply this method, it is necessary to develop prediction models with the capacity to capture the intrinsic differences between agricultural areas and incorporate them in the modeling process. High quality estimates are generally obtained when these models are applied to soil samples displaying characteristics similar to the samples used in their construction. However, low quality predictions are noted when applied to samples from new areas presenting different characteristics. One way to solve this problem is by recalibrating the models using selected samples from the area of interest. Based on this premise, the aim of this study was to use the spiking technique and spiking associated with hybridization to expand prediction models and estimate organic matter content in a target area undergoing different uses and management. A total of 425 soil samples were used for the generation of the state model, as well as 200 samples from a target area to select the subsets (10 samples) used for model recalibration. The spectral readings of the samples were obtained in the laboratory using the ASD FieldSpec 3 Jr. Sensor from 350 to 2500 nm. The spectral curves of the samples were then associated to the soil attributes by means of a partial least squares regression (PLSR). The state model obtained better results when recalibrated with samples selected through a cluster analysis. The use of hybrid spectral curves did not generate significant improvements, presenting estimates, in most cases, lower than the state model applied without recalibration. The use of the isolated spiking technique was more effective in comparison with the spiked and hybridized state models, reaching r2, square root of mean prediction error (RMSEP) and ratio of performance to deviation (RPD) values of 0.43, 4.4 g dm−3, and 1.36, respectively.

Soil contamination by potentially toxic elements (PTEs) is intensifying under increasing industrialization. Thus, the ability to efficiently delineate contaminated sites is crucial. Visible–near infrared (vis–NIR: 350–2500 nm) and X-ray fluorescence (XRF: 0.02–41.08 keV) spectroscopic techniques have attracted tremendous attention for the assessment of PTEs. Recently, the application of fused vis–NIR and XRF spectroscopy, which is based on the complementary effect of data fusion, is also increasing. Moreover, different data manipulation methods, including feature selection approaches, affect the prediction performance. This study investigated the feasibility of using single and fused vis–NIR and XRF spectra while exploring feature selection algorithms for the assessment of key soil PTEs. The soil samples were collected from one of the most heavily polluted areas of the Czech Republic and scanned using laboratory vis–NIR and XRF spectrometers. Univariate filter (UF) and genetic algorithm (GA) were used to select the bands of greater importance for the PTE prediction. Support vector machine (SVM) was then used to train the models using the full-range and feature-selected spectra of single sensors and their fusion. It was found that XRF spectra alone (primarily GA-selected) performed better than single vis–NIR and fused spectral data for predictions of PTEs. Moreover, the prediction models that were derived from the fused data set (particularly the GA-selected) enhanced the models’ accuracies as compared with the single vis–NIR spectra. In general, the results suggest that the GA-selected spectra obtained from the single XRF spectrometer (for As and Pb) and from the fusion of vis–NIR and XRF (for Pb) are promising for accurate quantitative estimation detection of the mentioned PTEs.

Multitemporal collections of satellite images and their products have recently been explored in digital soil mapping. This study aimed to produce a bare soil image (BSI) for the São Paulo State (Brazil) to perform a pedometric analysis for different geographical levels. First, we assessed the potential of the BSI for predicting the surface (0.00-0.20 m) and subsurface (0.80-1.00 m) clay, iron oxides (Fe 2 O 3 ), aluminum (m%) and bases saturation (V%) contents at the state level, which are important properties for soil classification. In this task, legacy soil samples, the BSI and terrain attributes were employed in machine learning. In a second moment, we evaluated the capacity of the BSI for clustering the landscape at the regional level, comparing the predicted patterns with a legacy semi-detailed soil map from a smaller reference site. In the final stage, the predicted soil maps from the state level were investigated at the farm level considering several sites distributed across the São Paulo state. Our results demonstrated that clay and Fe 2 O 3 reached the best prediction performance for both depths at the state level, reaching a RMSE of less than 10 %, RPIQ higher than 1.6 and R 2 of at least 0.41. Additionally, the predicted landscape clusters had a significant association with the main pedological classes, subsurface color, soil mineralogy and texture from the legacy semi-detailed soil map. Illustrative examples at the farm level indicated great capacity of BSI in detecting the variations of soils, which were linked to several soil properties, such as texture, iron content, drainage network, among others. Therefore, this study demonstrates that BSI is valuable information derived from optical Earth Observation data that can contribute to the future of soil survey and mapping in Brazil (PronaSolos).

Despite the natural spatial variability, cropland soils are subject to many interventions that can lead to alterations of soil functioning. As the cropland expansion took place in Brazil the last decades, leading to significant land-use change and environmental impacts, detailed information about soils is fundamental for sustainable development. Thus, considering the lack of spatially explicit information about cropland soils in Brazil, we aimed at performing high-resolution mapping of key topsoil attributes using spectral and terrain features extracted from Earth observation data (EOD). With the resulting information, we also aimed at performing a general examination of the main agricultural regions and estimate the total organic carbon stocks on croplands soils. For this, we obtained environmental predictors from the historical collection of Landsat data and the digital elevation model from Shuttle Radar Topographic Mission at the cloud-based platform of Google Earth Engine. The environmental predictors (30 m spatial resolution) with georeferenced soil samples (n = 5097) were used for predicting the topsoil content (0–20 cm) of clay, sand, silt, cation exchange capacity, pH, soil organic carbon (SOC) and SOC stock. Prediction models of clay, sand, SOC content, and SOC stocks had the best performance metrics, achieving a R2 ranging from 0.44 to 0.74 and ratio of performance to the interquartile range higher than 1.5. The predicted maps revealed the variability of topsoil among the cropped areas, indicating that the agricultural expansion took place on sandy soils. The SOC stock map provided consistent estimates compared to previous datasets but revealed additional information at the local and regional scales. Thus, this study supports the proposition that EOD is a valuable source for extracting environmental features for mapping and monitoring cropland soils at finer resolutions, assisting the evaluation of soil spatial distribution and the historical agriculture expansion over large geographical areas.

Geophysical methods, such as gamma-ray spectrometry, have great potential to enhance knowledge of the pedosphere (pedogenesis, pedogeochemistry and pedogeomorphology), helping to predict tropical soil attributes. We applied proximal gamma-ray spectrometry to evaluate tropical landscape dynamics, pedogenesis and spatial distribution of radionuclides and selected soil attributes. This study was carried out in southeast Brazil, where 79 soil samples (0–20 cm) along transects were collected to perform physical-chemical analysis coupled with collection of surface gamma-ray spectrometric data, allowing the detection of the radionuclides uranium (U238), thorium (Th232), and potassium (K40). Additionally, we analyzed soils in four toposequences, with varying lithology, relief, and hydrology. The radionuclide concentrations in soils showed a direct relationship with the parent material composition, either rocks or sediments. Weathering degree and the geochemical behavior of each radionuclide determines its permanence or removal in soils. Denudation processes along the toposequences also influence the distribution of radionuclides. On one hand, the radionuclide contents of mature, weathered soils are closely associated with advanced pedogenesis, with a higher clay contents and argic horizon formation, whereas in younger, less weathered soils the parent material exerts a greater influence than pedogenesis. Uranium decreased with altitude, and showed greater mobility compared with thorium. Thorium presented a higher correlation with clay content, and the opposite with sand content. Potassium detection increased with soil cation exchange capacity (CEC) and clay content. Gamma-spectrometry detected significant variations in some segments along the toposequences, undetectable by conventional soil survey techniques. This may indicate changes from one soil class to another or the continuity of a particular soil class, demonstrating the potential of this tool in digital soil mapping, pedogeochemical, pedogeomorphological and pedogenesis processes studies.

Weathering of the parent material is the initial process that plays an essential role in soil fertility. The study of this process has become crucial to better understand soil dynamics. Most studies on weathering are based on indexes that calculate the feldspar dissolution rate (for example, CIA, CIW, and PIA) which not always reflect the complexity in the soil profile. Other methodologies have been designed to improve these approaches, such as geochemical balance, difference of elemental contents (DE), and the W index. This work aim to assess the efficiency of these methods to identify variations in recent soil profiles developed from diabase as compared to the CIA, CIW, and PIA indexes. Data on the main oxides were used to calculate the geochemical balance, DE, CIA, CIW, PIA, and W indexes to analyze weathering intensity in two profiles originated from diabase in the São Paulo State, Brazil. The Ki and Kr indexes and Feo /Fed ratio were also factored and we analyzed the mineralogy of soil horizons and the source rock. The W index and DE presented greater weathering variation than the CIA, CIW, and PIA in both profiles. The geochemical balance, mineralogy, and the Ki, Kr and Feo /Fed ratios explained the variations between the horizons. The DE and geochemical balance together generated new information for the analysis of the weathering process than the isolated indexes. RESUMO O intemperismo do material de origem é o processo inicial que garantirá ao solo maior ou menor fertilidade, tornando o estudo desse processo fundamental para entender a dinâmica do solo. Grande parte dos estudos de intemperismo é baseada em índices que calculam a taxa de dissolução dos feldspatos (CIA, CIW e PIA, por exemplo) muitas vezes não refletindo a complexidade do intemperismo ocorrendo no perfil de solo. Outras metodologias- como balanço geoquímico, diferença elementar (DE) e índice W- têm sido desenhadas para melhorar esses estudos. Avaliar se essas metodologias serão mais eficientes que os índices CIA, CIW e PIA para apontar as variações intempéricas em perfis recentes desenvolvidos de basalto é o objetivo desse trabalho. Foram utilizados dados de óxidos principais para calcular o balanço geoquímico, DE, CIA, CIW, PIA e W e analisar a intensidade do intemperismo em dois perfis originados de basalto no Estado de São Paulo. Foram fatorados também os índices Ki e Kr e a relação Fe o/Fed e analisada a mineralogia dos horizontes dos solos e da rocha de origem. O índice W e a DE mostraram maior variação intempérica que CIA, CIW e PIA. O balanço geoquímico, a mineralogia e as razões Ki e Kr, Feo /Fed foram eficientes em explicar as variações que ocorreram entre os horizontes. A junção da DE com o balanço geoquímico trouxe novas informações para a análise do intemperismo que a observação isolada de CIA, CIW e PIA.

Simple Summary The whitefly Bemisia tabaci has become a primary pest in soybean fields in Brazil over the last decades, causing losses in the yield. Its reduced size and fast population growth make monitoring a challenge for growers. The use of hyperspectral proximal sensing (PS) is a tool that allows the identification of arthropod infested areas without contact with the plants. This optimizes the time spent on crop monitoring, which is important for large cultivation areas, such as soybean fields in Brazilian Cerrado. In this study, we investigated differences in the responses obtained from leaves of soybean plants, non-infested and infested with Bemisia tabaci in different levels, with the aim of its differentiation by using hyperspectral PS, which is based on the information from many contiguous wavelengths. Leaves were collected from soybean plants to obtain hyperspectral signatures in the laboratory. Hyperspectral curves of infested and non-infested leaves were differentiated with good accuracy by the responses of the bands related to photosynthesis and water content. These results can be helpful in improving the monitoring of Bemisia tabaci in the field, which is important in the decision-making of integrated pest management programs for this key pest. Abstract Although monitoring insect pest populations in the fields is essential in crop management, it is still a laborious and sometimes ineffective process. Imprecise decision-making in an integrated pest management program may lead to ineffective control in infested areas or the excessive use of insecticides. In addition, high infestation levels may diminish the photosynthetic activity of soybean, reducing their development and yield. Therefore, we proposed that levels of infested soybean areas could be identified and classified in a field using hyperspectral proximal sensing. Thus, the goals of this study were to investigate and discriminate the reflectance characteristics of soybean non-infested and infested with Bemisia tabaci using hyperspectral sensing data. Therefore, cages were placed over soybean plants in a commercial field and artificial whitefly infestations were created. Later, samples of infested and non-infested soybean leaves were collected and transported to the laboratory to obtain the hyperspectral curves. The results allowed us to discriminate the different levels of infestation and to separate healthy from whitefly infested soybean leaves based on their reflectance. In conclusion, these results show that hyperspectral sensing can potentially be used to monitor whitefly populations in soybean fields.

Geotechnologies allow natural resources to be surveyed more quickly and cheaply than traditional methods. This paper aimed to produce a digital soil map (DSM) based on Landsat time series data. The study area, located in the eastern part of the Brazilian Federal District (Rio Preto hydrographic basin), comprises a representative basin of the Central Brazil plateau in terms of pedodiversity. A spectral library was produced based on the soil spectroscopy (from the visible to shortwave infrared spectral range) of 42 soil samples from 0–15 cm depth using the Fieldspec Pro equipment in a laboratory. Pearson’s correlation and principal component analysis of the soil attributes revealed that the dataset could be grouped based on the texture content. Hierarchical clustering analysis allowed for the extraction of 13 reference spectra. We interpreted the spectra morphologically and resampled them to the Landsat 5 Thematic Mapper satellite bands. Afterward, we elaborated a synthetic soil/rock image (SySI) and a soil frequency image (number of times the bare soil was captured) from the Landsat time series (1984–2020) in the Google Earth Engine platform. Multiple Endmember Spectral Mixture Analysis (MESMA) was used to model the SySI, using the endmembers as the input and generating a DSM, which was validated by the Kappa index and the confusion matrix. MESMA successfully modeled 9 of the 13 endmembers: Dystric Rhodic Ferralsol (clayic); Dystric Rhodic Ferralsol (very clayic); Dystric Haplic Ferralsol (loam-clayic); Dystric Haplic Ferralsol (clayic); Dystric Petric Plinthosol (clayic); Dystric Petric Plinthosol (very clayic); Dystric Regosol (clayic); Dystric Regosol (very clayic); and Dystric, Haplic Cambisol (clayic). The root mean squared error (RMSE) varied from 0 to 1.3%. The accuracy of DSM achieved a Kappa index of 0.74, describing the methodology’s effectiveness to differentiate the studied soils.

Made available in DSpace on 2020-12-12T02:17:47Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-12-01 Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) The use of diffuse reflectance spectroscopy (DRS) has gained prominence in the quantification of soil attributes due to its ease and practicality of obtaining data. This study aimed to evaluate the potential of different methodologies applied to spectral curves given by DRS to estimate kaolinite (Kt) and gibbsite (Gb), and their spatial variability characterization for the Western Plateau of São Paulo. The Western Plateau of São Paulo has 13 million hectares, 2 million of them covered by basalt and 11 million by sandstones. A total of 600 samples were collected at a depth of 0.0–0.20 m. Calibration curves were constructed with pure minerals for x-ray diffraction (XRD) and DRS techniques. The Kt/(Kt + Gb) ratio and the percentages of Kt and Gb were determined by XRD and using the following three methodologies applied to spectral curves: continuum removal technique (CR), direct ratio of the valley (DRV), and multivariate analysis by partial least squares regression (PLSR). The CR procedure had means similar to those observed by XRD, i.e., 0.90 and 0.92, respectively, while DRV overestimated the ratio, with a mean of 1.32. DRS allowed the estimation of the Kt/(Kt + Gb) ratio for the different geological and landscape compartments of the Western Plateau of São Paulo for the CR and DRV procedures. CR procedure allowed constructing models to be more efficient compared to those obtained by DRV and PLSR. The use of geostatistics to interpolate the data of the ratio Kt/(Kt + Gb) by DRS provided important information to define specific management zones accurately and more economically. Department of Soils and Fertilizers Research Group CSME - Soil Characterization for Specific Management Faculty of Agrarian and Veterinary Sciences São Paulo State University (FCAV/UNESP) Department of Animal Science Faculty of Agrarian and Veterinary Sciences São Paulo State University (FCAV/UNESP) Departament of Soil Science Luiz de Queiroz College of Agriculture (ESALQ) University of São Paulo (USP) Departamento os Soil State University of Goiás (UEG) Department of Soils and Fertilizers Research Group CSME - Soil Characterization for Specific Management Faculty of Agrarian and Veterinary Sciences São Paulo State University (FCAV/UNESP) Department of Animal Science Faculty of Agrarian and Veterinary Sciences São Paulo State University (FCAV/UNESP)

Portable X-ray fluorescence (pXRF) has great potential for numerous applications in soil science. However, the basic knowledge about the effects of soil properties on pXRF spectra are still poorly studied, which may lead users to biased interpretations of mathematical models. The present study aimed to evaluate the outcomes of moisture, soil organic matter content (SOM), and iron forms on pXRF data. The work was conducted with seventeen soil samples from the central region of Sao Paulo state (Brazil). Three selective dissolution treatments were applied to remove: (i) soil organic matter (−SOM), ii) SOM and poorly crystalline iron forms (−o), iii) SOM and poorly crystalline plus well crystalline iron forms (−d). One additional treatment iv) including water addition (+W) was also carried out. The effects of treatments were evaluated for sandy and clayey samples. Soil particle size distribution and elemental content affected the bremsstrahlung and characteristic peaks counts. In +W, there was a generalized decrease in counts mainly for the light elements (magnesium, aluminum and silicon). Regarding the selective dissolution procedures, alterations were verified, reflecting mainly the removal power of reagents. Generally, the most pronounced alterations occurred for −d and moderate alterations for −SOM and −o. The pXRF data showed high correlation with particle size distribution and mineralogy attributes. The kaolinite, gibbsite, Fe2O3, Al2O3, SiO2, TiO2 and MnO contents were quantified with satisfactory accuracy (0.61

The regular monitoring of soil physical, chemical, and biological properties is very essential, due to its role in soil ecosystem functions. A cost-effective alternative for soil monitoring corresponds to spectral sensing techniques. Soil spectral sensing techniques can support decision-making in agricultural systems at both time and spatial scales, maximizing food production while preserving an adequate soil condition. Due to the large number of ground, airborne, and orbital spectral sensors operating today, this technology has been increasingly assimilated by soil scientists. However, it is important to have an adequate comprehension about the technique principles and limitations. This chapter provides a wide perspective about the soil spectral sensing in the visible (vis: 350–700 nm), near-infrared (NIR: 700–1000 nm), and shortwave infrared (SWIR: 1000–2500 nm), considering reflectance data at different acquisition levels. Here, it is discussed how soil constituents interact with EMR and the resulting soil spectral behaviors. We describe the predictive potential of vis-NIR-SWIR data for quantitative assessment of soil and which soil attributes have been reliably estimated and the most commonly used vis-NIR-SWIR equipment, as well as their advantages and limitations. Finally, we discuss the current application in soil science and future perspectives.

International audience; The soil mineralogical constitution directly influences its chemical, physical and hydraulic characteristics. Although very important, it is still rarely used for decision-making in agriculture, mainly due to the complexity and cost of standard analyzes. In this sense, the middle infrared spectroscopy (mid-IR, 4000 to 400 cm(-1)) has great potential to obtain soil mineralogical information quickly and accurately. Nevertheless, some soil constituents can severely influence the spectra and produce misinterpretations. In this research, we aim to detect changes in the mid-IR spectra caused by water, iron forms and organic matter (OM), and to relate soil attributes to laboratory spectra and remote sensing simulated spectral bands. The research area is located in Sao Paulo State, Brazil, where seventeen soil samples were collected. The reflectance intensities, shapes and absorption features of the mid-IR spectra before and after the removal of OM and iron forms and the addition of water were described. Soil attributes, such as kaolinite, gibbsite, 2:1 minerals among others were correlated with the mid-IR spectra and simulated ASTER spectral bands by Pearson's analysis, to verify its potential on mineralogical evaluation. The description of mid-IR revealed that the removal of the OM from the soil samples decreased the reflectance intensities between 4000 and 2000 cm(-1). Iron forms mainly influence the 3250 - 1200 cm(-1) spectral range and mask the spectral features of other minerals as well. The addition of water masked several absorption features and decreased the reflectance intensities from 3700 to 2700 cm(-1). High correlation coefficients were obtained between soil attributes and ASTER simulated spectral bands, which allowed the selection of potential spectral regions for future satellite sensors: 2760 - 2500 cm(-1) (3600 - 4000 nm), 2150 - 1875 cm(-1) (4600 - 5300 nm), and 840 - 740 cm(-1) (11900 - 3500 nm).

The importance of monitoring soil properties is constantly increasing among researchers and policy-makers. In this context,it is imperative to identify cost-effective and reliable strategies for soil mapping compared to the costlier traditionalsolutions. A wide range of tools are becoming available that enable better utilization of Earth Observation capabilities to monitor the soil ecosystem. This work is an effort of assessing the potential of Sentinel-2 imagery data for mapping SoilOrganic Matter (SOM) contents and investigating the possibilities of its enhancement through ASTER derived information. The rural area around the lake Zazari, located in the Western Macedonia district of Greece, was chosen as study area.Initially, pixel-wise vegetation indices (NDVI and NBR2) were calculated, utilizing a local version of the CEOS OpenData Cube for masking Sentinel-2 bare soil pixels extending a three-year period (2017–2019). The generated mask wasused to extract soil spectral signatures at the image level over selected 100 field samples. The resulting time series was expanded through the conjunction of ASTER Thermal InfraRed bands by matching the exact data acquisition dates of twoplatforms. The conclusive part of the work contains the application of regression modelling to effectively assess soil variables. The local Partial Least Square regression algorithm was chosen, due to its characteristics of performing inherently local predictions. Five-fold cross-validation technique was used for reporting the models’ accuracy, which was assessed through R2 coefficient, RPIQ ratio and RMSE. The model estimated SOM values among a synthetic bare soil composite image that was acquired over study area’s agricultural fields. Two models were trained and compared; one over Sentinel-2 imagery bands that were used as the predictor variables’ set and a second over an expanded predictor variables’ set, including ASTER thermal bands. The results signified evidence of accuracy increase of SOM content assessment, through spaceborne imagery analysis.

Terrain analysis is an important tool for modeling environmental systems. Aiming to use the cloud-based computing capabilities of Google Earth Engine (GEE), we customized an algorithm for calculating terrain attributes, such as slope, aspect, and curvatures, for different resolution and geographical extents. The calculation method is based on geometry and elevation values estimated within a 3 × 3 spheroidal window, and it does not rely on projected elevation data. Thus, partial derivatives of terrain are calculated considering the great circle distances of reference nodes of the topographic surface. The algorithm was developed using the JavaScript programming interface of the online code editor of GEE and can be loaded as a custom package. The algorithm also provides an additional feature for making the visualization of terrain maps with a dynamic legend scale, which is useful for mapping different extents: from local to global. We compared the consistency of the proposed method with an available but limited terrain analysis tool of GEE, which resulted in a correlation of 0.89 and 0.96 for aspect and slope over a near-global scale, respectively. In addition to this, we compared the slope, aspect, horizontal, and vertical curvature of a reference site (Mount Ararat) to their equivalent attributes estimated on the System for Automated Geospatial Analysis (SAGA), which achieved a correlation between 0.96 and 0.98. The visual correspondence of TAGEE and SAGA confirms its potential for terrain analysis. The proposed algorithm can be useful for making terrain analysis scalable and adapted to customized needs, benefiting from the high-performance interface of GEE.

The Earth’s surface dynamics provide essential information for guiding environmental and agricultural policies. Uncovered and unprotected surfaces experience several undesirable effects, which can affect soil ecosystem functions. We developed a technique to identify global bare surface areas and their dynamics based on multitemporal remote sensing images to aid the spatiotemporal evaluation of anthropic and natural phenomena. The bare Earth’s surface and its changes were recognized by Landsat image processing over a time range of 30 years using the Google Earth Engine platform. Two additional products were obtained with a similar technique: a) Earth’s bare surface frequency, which represents where and how many times a single pixel was detected as bare surface, based on Landsat series, and b) Earth’s bare soil tendency, which represents the tendency of bare surface to increase or decrease. This technique enabled the retrieval of bare surfaces on 32% of Earth’s total land area and on 95% of land when considering only agricultural areas. From a multitemporal perspective, the technique found a 2.8% increase in bare surfaces during the period on a global scale. However, the rate of soil exposure decreased by ~4.8% in the same period. The increase in bare surfaces shows that agricultural areas are increasing worldwide. The decreasing rate of soil exposure indicates that, unlike popular opinion, more soils have been covered due to the adoption of conservation agriculture practices, which may reduce soil degradation.

Studies on soil degradation are essential for environmental preservation. Since almost 30% of the global soils are degraded, it is important to study and map them for improving their management and use. We aimed to obtain a Soil Degradation Index (SDI) based on multi-temporal satellite images associated with climate variables, land use, terrain and soil attributes. The study was conducted in a 2598 km2 area in Sao Paulo State, Brazil, where 1562 soil samples (0–20 cm) were collected and analyzed by conventional methods. Spatial predictions of soil attributes such as clay, cation exchange capacity (CEC) and soil organic matter (OM) were performed using machine learning algorithms. A collection of 35-year Landsat images was used to obtain a multi-temporal bare soil image, whose spectral bands were used as soil attributes predictors. The maps of clay, CEC, climate variables, terrain attributes and land use were overlaid and the K-means clustering algorithm was applied to obtain five groups, which represented levels of soil degradation (classes from 1 to 5 representing very low to very high soil degradation). The SDI was validated using the predicted map of OM. The highest degradation level obtained in 15% of the area had the lowest OM content. Levels 1 and 4 of SDI were the most representative covering 24% and 23% of the area, respectively. Therefore, satellite images combined with environmental information significantly contributed to the SDI development, which supports decision-making on land use planning and management.

The demand for quality and low-cost soil information is growing due to the demands of land use planning and precision agriculture. Soil texture is one of the key soil properties, as it determines other vital soil characteristics such as soil structure, water and thermal regime, diversity of living organisms, plant growth, as well as the soil quality in general. It is usually not constant over an area, varying in space and with soil depth. Routine soil texture analysis is, however, time consuming and expensive. Because of this, the success of proximal soil sensing techniques in estimate soil properties using the VIS-NIR-SWIR and MIR regions is increasing. Advantages of soil spectroscopy include time efficiency, economic convenience, non-destructive application and freeing of chemical agents involved. Therefore, the objectives of this study were: (a) to explore the potential of clay, sand and silt prediction using reflectance spectroscopy; (b) assess the performance of predictive models in different spectral regions, i.e. VIS-NIR-SWIR and MIR; (c) assess the effect of different soil depths on predictive models; and finally (d) explain the differences in prediction accuracy in the means of the input data structure. Soil samples were collected at three depths (0–20, 20–40 and 40–60 cm) at 70 sampling sites over a study area located in the State of Rio Grande do Sul (Brazil). The content of soil texture was determined by Pipette method, and soil spectra were obtained with FieldSpec Pro (VIS-NIR-SWIR) and by Alpha Sample Compartment RT (MIR). Cubist regression algorithm was applied to train predictive models in three separate modeling modes differing in spectral region: (i) VIS-NIR-SWIR, (ii) MIR and (iii) VIS-NIR-SWIR plus MIR. The results showed that the combination of all three soil depths led to a more accurate prediction of soil texture compared to subdivided soil depths. This was explained by variability of the data, which was larger for the total dataset than for the depth-specific data. Consequently, we suggested that no precise comparison between different studies can be made without a proper description of the input data. For all-depths models, the MIR calibration obtained the best accuracy, which was explained due to more information comprised in the MIR region against the VIS-NIR-SWIR. The bands that were more important in predicting soil texture in MIR are related to mineralogy, specifically to kaolinite. This study demonstrated that the MIR spectroscopy technique is capable to complement the standard soil particle size analysis, specially where a large number of soil samples need to be treated in a short period of time.

The number of samples used in the calibration data set affects the quality of the generated predictive models using visible, near and shortwave infrared (VIS–NIR–SWIR) spectroscopy for soil attributes. Recently, the convolutional neural network (CNN) has been regarded as a highly accurate model for predicting soil properties on a large database. However, it has not yet been ascertained how large the sample size should be for CNN model to be effective. This paper investigates the effect of the training sample size on the accuracy of deep learning and machine learning models. It aims at providing an estimate of how many calibration samples are needed to improve the model performance of soil properties predictions with CNN as compared to conventional machine learning models. In addition, this paper also looks at a way to interpret the CNN models, which are commonly labelled as a black box. It is hypothesised that the performance of machine learning models will increase with an increasing number of training samples, but it will plateau when it reaches a certain number, while the performance of CNN will keep improving. The performances of two machine learning models (partial least squares regression – PLSR; Cubist) are compared against the CNN model. A VIS–NIR–SWIR spectra library from Brazil, containing 4251 unique sites with averages of two to three samples per depth (a total of 12 044 samples), was divided into calibration (3188 sites) and validation (1063 sites) sets. A subset of the calibration data set was then created to represent a smaller calibration data set ranging from 125, 300, 500, 1000, 1500, 2000, 2500 and 2700 unique sites, which is equivalent to a sample size of approximately 350, 840, 1400, 2800, 4200, 5600, 7000 and 7650. All three models (PLSR, Cubist and CNN) were generated for each sample size of the unique sites for the prediction of five different soil properties, i.e. cation exchange capacity, organic carbon, sand, silt and clay content. These calibration subset sampling processes and modelling were repeated 10 times to provide a better representation of the model performances. Learning curves showed that the accuracy increased with an increasing number of training samples. At a lower number of samples (< 1000), PLSR and Cubist performed better than CNN. The performance of CNN outweighed the PLSR and Cubist model at a sample size of 1500 and 1800, respectively. It can be recommended that deep learning is most efficient for spectra modelling for sample sizes above 2000. The accuracy of the PLSR and Cubist model seems to reach a plateau above sample sizes of 4200 and 5000, respectively, while the accuracy of CNN has not plateaued. A sensitivity analysis of the CNN model demonstrated its ability to determine important wavelengths region that affected the predictions of various soil attributes.

Knowing particularities of soils allows the adoption of sustainable management practices. The most efficient method to obtain data on the soil is through its characterization, essential for planning land use and soil conservation. However, soil surveys conducted in the conventional methods are costly and time-consuming. This study aimed to identify spectral patterns in the mid-IR wavelengths for a quick and low cost characterization of soil classes. Soils from the Natuba river basin, Pernambuco State, Brazil, were used. These soils were mapped on a scale of 1:25,000, where the presence of Latossolos, Argissolos, Gleissolos and Neossolos were identified. The data of reflectance for profiles of each soil class were collected using the spectral range between 2,500 and 25,000 nm. In the mid-IR region, Latossolos and Argissolos presented spectral characteristics peculiar to classification criteria. The increased contents of organic matter and iron oxides reduced soil reflectance. The horizons with sand content above 80% showed strong absorption spectra and significant reflectance peaks of quartz in the mid-IR. The wavelengths 2,681, 2,600 and 2,495 cm-1 occurred only in these horizons. The spectral analysis presented as a high-potential method for the characterization and classification of soils. Made available in DSpace on 2021-11-11T18:01:23Z (GMT). No. of bitstreams: 1 Spectral-patterns-in-the-mid-ir-of-soils-2020.pdf: 1946587 bytes, checksum: accd3d9898b084c2f681d811735bd48b (MD5) Previous issue date: 2020

Reflectance of light across the visible, near-infrared and shortwave infrared (VIS-NIR-SWIR, 0.4–2.5 µm) spectral region is very useful for investigating mineralogical, physical and chemical properties of soils, which can reduce the need for traditional wet chemistry analyses. As many collections of multispectral satellite data are available for environmental studies, a large extent with medium resolution mapping could be benefited from the spectral measurements made from remote sensors. In this paper, we explored the use of bare soil composites generated from the large historical collections of Landsat images for mapping cropland topsoil attributes across the European extent. For this task, we used the Geospatial Soil Sensing System (GEOS3) for generating two bare soil composites of 30 m resolution (named synthetic soil images, SYSI), which were employed to represent the median topsoil reflectance of bare fields. The first (framed SYSI) was made with multitemporal images (2006–2012) framed to the survey time of the Land-Use/Land-Cover Area Frame Survey (LUCAS) soil dataset (2009), seeking to be more compatible to the soil condition upon the sampling campaign. The second (full SYSI) was generated from the full collection of Landsat images (1982–2018), which although displaced to the field survey, yields a higher proportion of bare areas for soil mapping. For evaluating the two SYSIs, we used the laboratory spectral data as a reference of topsoil reflectance to calculate the Spearman correlation coefficient. Furthermore, both SYSIs employed machine learning for calibrating prediction models of clay, sand, soil organic carbon (SOC), calcium carbonates (CaCO3), cation exchange capacity (CEC), and pH determined in water, using the gradient boosting regression algorithm. The original LUCAS laboratory spectra and a version of the data resampled to the Landsat multispectral bands were also used as reference of prediction performance using VIS-NIR-SWIR multispectral data. Our results suggest that generating a bare soil composite displaced to the survey time of soil observations did not improve the quality of topsoil reflectance, and consequently, the prediction performance of soil attributes. Despite the lower spectral resolution and the variability of soils in Europe, a SYSI calculated from the full collection of Landsat images can be employed for topsoil prediction of clay and CaCO3 contents with a moderate performance (testing R2, root mean square error (RMSE) and ratio of performance to interquartile range (RPIQ) of 0.44, 9.59, 1.77, and 0.36, 13.99, 1.54, respectively). Thus, this study shows that although there exist some constraints due to the spatial and temporal variation of soil exposures and among the Landsat sensors, it is possible to use bare soil composites for mapping key soil attributes of croplands across the European extent.

Soil texture and organic carbon (OC) content influence the spectral response. These attributes are relevant for the preservation and proper management of land use in the pursuit of a sustainable agriculture. Laboratory and satellite sensors have been applied as a powerful tool for studying so is, but their analysis using these sensors has mainly focused on the visible (Vis), near infrared (NIR) and shortwave infrared (SWIR) regions of the electromagnetic spectrum, with few studies in the Medium Infrared (MIR). The aim of this study was to identify the spectral pattern of soils with different granulometry (sand and clay) and OC content using laboratory and satellite sensors in the MIR region, specifically in the Thermal Infrared (TIR) range (ASTER, Landsat satellites). The study performed qualitative and quantitative analyses of clay, OC and sand fractions (fine and coarse). The study area is located in the region of Piracicaba, Sao Paulo, Brazil, where collected 150 soil samples (0–20 cm depth). Soil texture was determined by the pipette method and OC via dry combustion. Reflectance and emissivity (Ɛ) spectral data were obtained with the Fourier Transform Infrared (FT-IR) Alpha sensor (Bruker Optics Corporation). An image “ASTER_05” from July 15, 2017 was acquired with values of Ɛ. Samples were separated by textural classes and the spectral behavior in the TIR region was described. The data obtained by the laboratory sensor were resampled to the satellite sensor bands. The behavior between spectra of both sensors was similar and had significant correlation with the studied attributes, mainly sand. For the partial least squares regression (PLSR) models, six strategies were used (MIR, MIR_ASTER, ASTER, TIR, TIR Correlation Index (TIR_CID), and MIR Correlation Index (MIR_CID)), which consisted in the use of all sensors bands, or by the selection of bands that presented the most significant correlations with each one of the attributes. Models presented a good performance in the prediction of all attributes using the whole MIR. In the TIR region, the models for total sand content and for fine and coarse fractions were good. Models created with ASTER sensor data were not as promising as those with laboratory ones. The use of specific bands was useful in estimating some attributes in the MIR and TIR, improving the predictive performance and validation of models. Therefore, the discrimination of soil attributes with satellite sensors can be improved with the identification of specific bands, as observed in the results with laboratory sensors.

The soil carbon pool which is the sum of soil organic carbon (SOC) and soil inorganic carbon (SIC) is the second largest active store of carbon after the oceans and it is an important component of the global carbon cycle. Hence, accurate estimation of SOC and SIC as important carbon reservoirs in terrestrial ecosystems using fast, inexpensive and non-destructive methods is crucial for planning different climate change policies. The aim of the current research was to examine the effectiveness of Vis-NIR (visible and near-infrared spectroscopy: 350–2500 nm) and MIR (mid-infrared spectroscopy: 4000–400 cm−1) to characterize and estimate soil organic matter (SOM) and carbonates as main components of soil carbon stocks in Juneqan, Charmahal va Bakhtiari, Iran. To do so, a total of 548 soil samples from this area were collected (October 2015) and analyzed in laboratory (August 2017). In order to develop models capable of predicting SOM and carbonates content, seven spectral preprocessing methods comprising Absorbance (Abs), De-trending (Det), Continuum removal (CR), Savitzky-Golay derivatives (SGD), standard normal variate transformation (SNV), multiplicative scatter correction (MSC) and Normalization by range (NBR) were conducted along with five multivariate methods including Random Forest (RF), Partial Least-Squares Regression (PLSR), Artificial Neural Network (ANN), Support Vector Machine (SVM) and Gaussian Process Regression (GPR). The content of carbonates caused spectral reflectance intensity to augment on several ranges of spectrum and strong absorption feature at 2338 nm in the Vis-NIR and 714, 850, 870, 1796, 2150 and 2510 cm−1 in the MIR spectra range. SOM absorbed energy in several ranges, but also showed specific peaks in MIR. Both facts are associated with the structure of carbonates and SOM and its interaction with energy. The best combination of preprocessing and calibration models for carbonates quantification in Vis-NIR spectra was Det/PLSR (R2= 0.74, RPD= 2.19, RMSE= 6.45). For SOM, it was Det/PLSR (R2= 0.82, RPD= 2.41, RMSE= 0.75). The Det/RF (R2= 0.87, RPD= 2.44, RMSE= 0.66) for the quantification of SOM and MSC/RF (R2= 0.84, RPD= 2.84, RMSE= 5.50) for carbonates in MIR spectra range showed the greatest results. The stronger occurrence of spectral bands in MIR as well as the specificity of the absorption features indicated that this range produced better predictions. The obtained results highlighted the significant role of soil spectroscopy technique in predicting SOC and soil carbonates as key components of soil carbon stocks in the study area. Therefore, this technique can be used as a more cost-effective, time saving and nondestructive alternative to traditional methods of soil analysis.

Most studies assessing land-use change (LUC) and management impacts on the soil organic carbon (SOC) budget mainly focused on total soil organic matter (SOM), neglecting its functional fractions such as water-extractable organic matter (WEOM). This may limit a better understanding of SOC dynamics along the soil profile as WEOM plays a key role in several soil and microbial processes. Here, we aimed to obtain a quantitative and qualitative assessment of WEOM dynamics to 1 m depth in areas after native vegetation conversion to long-term agriculture in Brazil. For this purpose, cropland areas under no-tillage (NT; ∼ 23 years old) and native vegetation (NV; i.e., Cerrado vegetation) were evaluated across a transect of 1000 km within Brazil’s new agricultural frontier – the so-called MATOPIBA region. We combined spectroscopic analyses (UV–Vis and DRIFT spectroscopies), C and nitrogen (N) determination in different SOM pools (i.e., bulk SOM, WEOM, and microbial biomass), and microbial measurements [e.g., microbial biomass C (MBC) and N (MBN), and β-glucosidase activity] to better understand changes in WEOM dynamics induced by LUC. We observed that although SOC decreased along the soil profile after NV (4.2–20.7 g kg−1) conversion to NT (3.8–14.2 g kg−1), water-extractable organic C levels (3.6–79.3 mg L−1) were similar between land uses. On the other hand, an increase in water-extractable total N was observed after NV (0.4–4.6 mg L−1) conversion to NT (0.5–19.3 mg L−1). Although MBC and MBN levels decreased for most study sites, β-glucosidase activity increased after LUC indicating intensive microbial processing of SOM. As a result, WEOM had lower aromaticity and molecular weight (i.e., low values of SUVA254 and A2/A3) in NT than NV; and a relative decomposition of aliphatic and polysaccharides over aromatic and amine/amide was observed after LUC through DRIFT spectroscopy. Our study indicates that LUC had stronger impacts on WEOM quality than its quantity, with agricultural areas having more biochemical changes in WEOM than NV.

Soil contamination by potentially toxic elements (PTEs) is one of the greatest threats to environmental degradation. Knowing where PTEs accumulated in soil can mitigate their adverse effects on plants, animals, and human health. We evaluated the potential of using long-term remote sensing images that reveal the bare soils, to detect and map PTEs in agricultural fields. In this study, 360 soil samples were collected at the superficial layer (0–20 cm) in a 2574 km2 agricultural area located in Sao Paulo State, Brazil. We tested the Soil Synthetic Image (SYSI) using Landsat TM/ETM/ETM+, Landsat OLI, and Sentinel 2 images. The three products have different spectral, temporal, and spatial resolutions. The time series multispectral images were used to reveal areas with bare soil and their spectra were used as predictors of soil chromium, iron, nickel, and zinc contents. We observed a strong linear relationship (−0.26 > r > −0.62) between the selected PTEs and the near infrared (NIR) and shortwave infrared (SWIR) bands of Sentinel (ensemble of 4 years of data), Landsat TM (35 years data), and Landsat OLI (4 years data). The clearest discrimination of soil PTEs was obtained from SYSI using a long term Landsat 5 collection over 35 years. Satellite data could efficiently detect the contents of PTEs in soils due to their relation with soil attributes and parent materials. Therefore, distinct satellite sensors could map the PTEs on tropics and assist in understanding their spatial dynamics and environmental effects.