Search

Soil spectroscopy is a widely used method for estimating soil properties that are important to environmental and agricultural monitoring. However, a bottleneck to its more widespread adoption is the need for establishing large reference datasets for training machine learning (ML) models, which are called soil spectral libraries (SSLs). Similarly, the prediction capacity of new samples is also subject to the number and diversity of soil types and conditions represented in the SSLs. To help bridge this gap and enable hundreds of stakeholders to collect more affordable soil data by leveraging a centralized open resource, the Soil Spectroscopy for Global Good initiative has created the Open Soil Spectral Library (OSSL). In this paper, we describe the procedures for collecting and harmonizing several SSLs that are incorporated into the OSSL, followed by exploratory analysis and predictive modeling. The results of 10-fold cross-validation with refitting show that, in general, mid-infrared (MIR)-based models are significantly more accurate than visible and near-infrared (VisNIR) or near-infrared (NIR) models. From independent model evaluation, we found that Cubist comes out as the best-performing ML algorithm for the calibration and delivery of reliable outputs (prediction uncertainty and representation flag). Although many soil properties are well predicted, total sulfur, extractable sodium, and electrical conductivity performed poorly in all spectral regions, with some other extractable nutrients and physical soil properties also performing poorly in one or two spectral regions (VisNIR or NIR). Hence, the use of predictive models based solely on spectral variations has limitations. This study also presents and discusses several other open resources that were developed from the OSSL, aspects of opening data, current limitations, and future development. With this genuinely open science project, we hope that OSSL becomes a driver of the soil spectroscopy community to accelerate the pace of scientific discovery and innovation. [ABSTRACT FROM AUTHOR]

• Large differences in relative absorbances were found across twenty instruments. • All instruments delivered good predictions when calibrated internally. • Predictive performance varied when a single existing library was used for calibration. • Standard Normal Variate helps reduce dissimilarities across instruments. • However, spectral standardization may be necessary in contrasting situations. Diffuse reflectance spectroscopy has been extensively employed to deliver timely and cost-effective predictions of a number of soil properties. However, although several soil spectral laboratories have been established worldwide, the distinct characteristics of instruments and operations still hamper further integration and interoperability across mid-infrared (MIR) soil spectral libraries. In this study, we conducted a large-scale ring trial experiment to understand the lab-to-lab variability of multiple MIR instruments. By developing a systematic evaluation of different mathematical treatments with modeling algorithms, including regular preprocessing and spectral standardization, we quantified and evaluated instruments' dissimilarity and how this impacts internal and shared model performance. We found that all instruments delivered good predictions when calibrated internally using the same instruments' characteristics and standard operating procedures by solely relying on regular spectral preprocessing that accounts for light scattering and multiplicative/additive effects, e.g., using standard normal variate (SNV). When performing model transfer from a large public library (the USDA NSSC-KSSL MIR library) to secondary instruments, good performance was also achieved by regular preprocessing (e.g., SNV) if both instruments shared the same manufacturer. However, significant differences between the KSSL MIR library and contrasting ring trial instruments responses were evident and confirmed by a semi-unsupervised spectral clustering. For heavily contrasting setups, spectral standardization was necessary before transferring prediction models. Non-linear model types like Cubist and memory-based learning delivered more precise estimates because they seemed to be less sensitive to spectral variations than global partial least square regression. In summary, the results from this study can assist new laboratories in building spectroscopy capacity utilizing existing MIR spectral libraries and support the recent global efforts to make soil spectroscopy universally accessible with centralized or shared operating procedures. [ABSTRACT FROM AUTHOR]

• Spectral and terrain features for mapping cropland soils in Brazil. • Detailed soil spatial distribution and SOC stock was estimated in croplands. • Predicted maps revealed a high variability of topsoil among the cropped areas. • Earth Observation data is a valuable source for mapping and monitoring cropland soils. 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. [ABSTRACT FROM AUTHOR]

The Midwest region in Brazil has the largest and most recent agricultural frontier in the country where there is no currently detailed soil information to support the agricultural intensification. Producing large-extent digital soil maps demands a huge volume of data and high computing capacity. This paper proposed mapping surface and subsurface key soil attributes with 30 m-resolution in a large area of Midwest Brazil. These soil maps at multiple depth increments will provide adequate information to guide land use throughout the region. The study area comprises about 851,000 km2 in the Cerrado biome (savannah) in the Brazilian Midwest. We used soil data from 7908 sites of the Brazilian Soil Spectral Library and 231 of the Free Brazilian Repository for Open Soil Data. We selected nine key soil attributes for mapping and aggregated them into three depth intervals: 0–20, 20–60 and 60–100 cm. A total of 33 soil predictors were prepared using Google Earth Engine (GEE), such as climate and geologic features with 1 km-resolution, terrain and two new covariates with 30 m-resolution, based on satellite measurements of the topsoil reflectance and the seasonal variability in vegetation spectra. The scorpan model was adopted for mapping of soil variables using random forest regression (RF). We used the model-based optimization by tuning RF hyperparameters and calculated the scaled permutation importance of covariates in R software. Our results were promising, with a satisfactory model performance for physical and chemical attributes at all depth intervals. Elevation, climate and topsoil reflectance were the most important covariates in predicting sand, clay and silt. In general, for predicting soil chemical attributes, climatic variables, elevation and vegetation reflectance provided to be the most important of predictive components, while for organic matter it was a combination of climatic dynamics and reflectance bands from vegetation and topsoil. The multiple depth maps showed that soil attributes largely varied across the study area, from clayey to sandy, suggesting that less than 44% of the studied soils had good natural fertility. We concluded that key soil attributes from multiple depth increments can be mapped using Earth observations data and machine learning methods with good performance. [ABSTRACT FROM AUTHOR]

Pedological assessment and mapping by hyperspectral proximal and multispectral remote sensing comes up as an important alternative for large extent areas with high soil variability. Strategies indicating how to integrate these multi-source data for digital soil mapping are emerging. In this paper, we used proximal and remote sensed data to perform pedological assessments and to support a detailed soil class mapping by MESMA. The study area is located in Southeast of Federal District, Brazil. Six toposequences were defined according to pedomorphogeological assessments and 34 sites were collected for laboratory analysis (texture and chemical) and VIS-NIR-SWIR (350–2500 nm) spectroscopy reflectance analysis. We assessed soil mineralogy based on derivative analysis of topsoil reflectance and we grouped observations to recognize topsoil spectral patterns by clustering method. Topsoil patterns (mean spectral curve for each cluster) were convolved using a Gaussian function of Landsat 5-TM spectral bands to obtain endmembers. Then, we used a Landsat 5 TM time series to produce a bare soil composite denominated Synthetic Soil Image (SYSI). Endmembers and SYSI were used as input data for Multiple Endmember Spectral Mixture Analysis (MESMA) to map the soil classes. Topsoil spectra clustered soil samples that were similar in texture, mineralogy and color, and identified 13 topsoil patterns (endmembers). SYSI retrieved 74% of bare soil area coverage and presented very similar spectral patterns to endmembers. The RGB 543 composite highlighted the mineralogy (i.g. sesquioxides, kaolinite and quartz), texture and color of the soils. MESMA modeled almost 100% of SYSI from endmembers, with low global RMSE of 0.86% and high global fraction of 62%. Soil classes were mapped using topsoil reflectance patterns and satellite images by MESMA method. We validated the digital soil map by using independent field-visited sites, which reached a Kappa coefficient of 0.73. We efficiently assessed soil mineralogy and recognized patterns from laboratory topsoil spectra. We successfully mapped soil classes using topsoil reflectance patterns and a bare soil composite image by MESMA method. Flowchart showing the procedure applied to produce the digital soil map where the soils are classified up to the second categorical level (subgroup) of the Brazilian Soil Classifications System (SiBCS). Unlabelled Image • Pedological mapping based on topsoil spectra is feasible. • We assessed soil mineralogy and recognized patterns from laboratory topsoil spectra. • We mapped soil classes using topsoil spectra and a bare soil composite by MESMA. • The non-detection of subsurface reflectance was not impeditive for soil mapping. • MESMA-derived digital soil map reached Kappa of 0.73. [ABSTRACT FROM AUTHOR]

Abstract 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 São 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 R2 adj. The correlation between the spectra of both depths was significant and the soil attributes prediction reached an R2 adj 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. Highlights • Subsurface soil attributes can be estimated by spectral pedotransfer functions • Bare soil composites support spectral pedotransfer functions • Bare soil composites assist in digital soil mapping [ABSTRACT FROM AUTHOR]

There is a consensus about the necessity to achieve a quick soil spatial information with few human resources. Remote/proximal sensing and pedotransference are methods that can be integrated into this approach. On the other hand, there is still a lack of strategies indicating on how to put this in practice, especially in the tropics. Thus, the objective of this work was to suggest a strategy for the spatial prediction of soil classes by using soil spectroscopy from ground laboratory spectra to space images platform, as associated with terrain attributes and spectral libraries. The study area is located in São Paulo State, Brazil, which was covered by a regular grid (one per ha), with 473 boreholes collected at top and undersurface. All soil samples were analyzed in laboratory (granulometry and chemical), and scanned with a VIS-NIR-SWIR (400–2500 nm) spectroradiometer. We developed two traditional pedological maps with different detail levels for comparison: TFS-1 regarding orders and subgroups; and TFS-2 with additional information such as color, iron and fertility. Afterwards, we performed a digital soil map, generated by models, which used the following information: (i) predicted soil attributes from undersurface layer (diagnostic horizon), obtained by using a local spectral library; (ii) spectral reflectance of a bare soil surface obtained by Landsat image; and (iii) derivative of terrain attributes. Thus, the digital map was generated by a combination of three variables: remote sensing (Landsat data), proximal sensing (laboratory spectroscopy) and relief. Landsat image with bare soil was used as a first observation of surface. This strategy assisted on the location of topossequences to achieve soil variation in the area. Afterwards, spectral undersurface information from these locations was used to modelize soil attributes quantification (156 samples). The model was used to quantify samples in the entire area by spectra (other 317 samples). Since the surface was bare soil, it was sampled by image spectroscopy. Indeed, topsoil spectral laboratory information presented great similarity with satellite spectra. We observed angle variation on spectra from clayey to sandy soils as differentiated by intensity. Soil lines between bands 3/4 and 5/7 were helpful on the link between laboratory and satellite data. The spectral models of soil attributes (i.e., clay, sand, and iron) presented a high predictive performance (R2 0.71 to 0.90) with low error. The spatial prediction of these attributes also presented a high performance (validations with R2 > 0.78). The models increased spatial resolution of soil weathering information using a known spectral library. Elevation (altitude) improved mapping due to correlation with soil attributes (i.e., clay, iron and chemistry). We observed a close relationship between soil weathering index map and laboratory spectra + image spectra + relief parameters. The color composite of the 5R, 4G and 3B had great performance on the detection of soils along topossequences, since colors went from dark blue to light purple, and were related with soil texture and mineralogy of the region. The comparison between the traditional and digital soil maps showed a global accuracy of 69% for the TFS-1 map and 62% in the TFS-2, with kappa indices of 0.52 and 0.45, respectively. We randomly validated both digital and traditional maps with individual plots at field. We achieve a 75% and 80% agreement for digital and traditional maps, respectively, which allows us to conclude that traditional map is not necessarily the truth and digital is very close. The key of the strategy was to use bare soil image as a first step on the indication of soil variation in the area, indicating in-situ location for sample collection in all depths. The current strategy is innovative since we linked sensors from ground to space in addition with relief parameters and spectral libraries. The strategy indicates a more accurate map with less soil samples and lower cost. [ABSTRACT FROM AUTHOR]

• Hydrological data contributed more than relief for prediction of sand, clay, and OM. • Confluence angle and channel sinuosity have distinct patterns over different soil types. • Hydrology, relief, and SySI data resulted in R2 of 0.71 for clay prediction. • Drainage morphometry changes across the landscape according to the soil class. Drainage network (DN) is the representation of all the stream channels developed over the landscape. The morphometry of DN describes the relationship between channel characteristics and basin geometry, which is regulated by a series of processes, such as weathering, geomorphology, sediment erosion/deposition. The interaction between these factors impacts soil formation, resulting in a relationship between DN morphometry and soil characteristics. Unstable surfaces produced shallow soils, enhancing surface runoff and high drainage density, while stable landscapes favor vertical infiltration through old and weathered soils. Digital soil mapping has benefited from multiple environmental variables, such as relief and satellite data, but DN information can offer great contributions for the prediction of soil attributes. In this work, we applied a set of complex DN variables to perform digital maps of clay, sand, and soil organic carbon (SOC), for a 1378 km2 site in the São Paulo state, Brazil. We analyzed the relationship between the drainage density (DD), drainage frequency (DF), channel sinuosity, and confluence angle with soil classes and attributes. The Cubist and Random Forest algorithms were tested to predict the soil information, and to evaluate the impact of the new hydrological variables. The results showed that landscapes predominated by clayey soils favor surface runoff and increase channel formation, with higher channel sinuosity and acute confluence angles. These new drainage variables contributed 35 to 55% for SOC prediction. DD and DF were the most important drainage variables on the models, ranging from 65 to 70%. The external validation reached R2 of 0.72 and 0.56 for the prediction of clay and sand, respectively. The impact of DN information on the model performance suggests that more work is needed to better explore and understand the relationship between DN and soil information. [ABSTRACT FROM AUTHOR]

Parent material is one of the five factors in soil formation. Studies on parent material allow interpreting soil genesis processes and improve our knowledge of specific soil attributes. However, soil parent material maps at detailed cartographic scale (finer than 1:100,000) are rare in tropical areas and it is usually inferred from poorly detailed geological data, which generally group different lithologies into single units. Thus, we propose a methodology to map soil parent material based on remote sensing and machine learning in a geologically very complex area. The study site covers 1378 km2 in São Paulo State, Brazil. Prediction models used data from 280 geological observation points, a digital elevation model (spatial resolution of 5 m, upscale to 30 m) and multitemporal Landsat images in a range of 30 years. We evaluated six classification algorithms, namely random forest, decision tree, support vector machine, multinomial logistic regression, K-means (unsupervised classification), and object-based image analysis with maximum likelihood classification. Environmental covariates were grouped to create different scenarios combining terrain derivatives, hydrologic covariates, topsoil spectral reflectance, and spatial coordinates. A bare soil image, elaborated using 30 years of Landsat data, was evaluated as a covariate to predict soil parent material. Predictions were validated using three different strategies: cross-validation, separate validation dataset (20%), and comparison with legacy geological maps (information from two areas with geological maps at fine scale). We also assessed the correspondence between the map of predicted soil parent material and data of soil particle size from 571 soil sampling points. Random forest algorithm presented the best validation performance, whereas the group of terrain derivatives and hydrologic covariates explained most of model variation. The produced parent material map was coherent with the spatial distribution of soil particle size across the study area. Unlabelled Image • A soil parent predicted showed details not present in an available geological map. • Random Forest showed the highest accuracy and Kappa coefficient. • Spatial coordinates were important covariates. However, may cause artifacts. • The predicted map showed high correspondence with soil particle size distribution. [ABSTRACT FROM AUTHOR]

• Lithology, pedogenesis and relief affect soil magnetic susceptibility. • Pedogenesis strongly affect magnetic susceptibility on more evolved soil classes. • Lithology contribute more to magnetic susceptibility in less evolved soil classes. • Magnetic susceptibility had good correlation with soil texture, CEC and iron content. • Field sensor can be used in delimitation of soil classes and digital soil mapping. Soil magnetic susceptibility (κ) has potential to be used as a pedoenvironmental indicator from which mineralogy, pedogeochemical, pedogeomorphological and pedogenic processes can be inferred. It can be used in pedosphere studies, as an auxiliary information for appropriate and sustainable soil use and management. This research aimed to analyze how pedogenesis and geochemical processes affect the κ and some of its attributes, as well as its potential use in discriminating soil great groups, following the digital soil mapping approach. The study area is located in São Paulo State - Brazil. Soil samples were collected for physical–chemical analysis from 79 locations (0–20 cm depth). At these sites, magnetic susceptibility was measured with a portable field instrument and analyzed in terms of geology, relief and soil class. The results showed that geology strongly affects κ, mainly in diabase derived soils, followed by metamorphosed siltstone and siltstone. In fluvial sediments, the κ exhibits different behaviors due to different sediments deposited by the Capivari River. In less evolved soils, such as Cambisols, lithology is a more important contributor to κ than pedogenesis. In more evolved soils, pedogenesis increases κ, whereas argilluviation/ferralitization reduces it. The κ values did not decrease significantly or even increase downslope, due to the presence of diabase on the lower parts. Differences in κ where observed between diabase bedrock located in different parts of the study area, indicating more of an influence by geomorphic processes rather than lithology. With respect to soil attributes, positive correlations between κ and base saturation, cation exchange capacity, organic matter, and iron and clay content were found, whereas a negative correlation was found between κ and sand content. The κ correlates with changes in lithology and soil class demonstrating its application as a potential tool for the discrimination of soil great groups and digital soil mapping. [ABSTRACT FROM AUTHOR]

Textural differentiation between soil horizons occurs due to the soil formation processes and is one of the main characteristics used for soil classification and erosion risk analysis. The aim of this work was to develop a Ratio of Clay Spectroscopic Index (RCSI) and establish a degree of granulometric differentiation between soil horizons using VIS-NIR-SWIR (350–2500 nm) and MIR (2500–25,000 nm) data. We described morphologically and collected 150 soil profile samples of five representative soils from the southeastern and central-western Brazil. Three spectral models (VIS-NIR-SWIR, MIR and VIS-NIR-SWIR-MIR) were classified in terms of analytical organic matter contents from A and B horizons. We constructed 15 RCSI and considered reflectance difference for specific bands or band combinations, which were selected based on the qualitative and quantitative spectral signatures assessment. The soil horizons were differentiated through their spectral regions associated to iron oxides, clay minerals and quartz. The use of RCSI-15 resulted in a R2 of 0.79 and a RPD of 2.21. The indices reduce the dependence on advanced statistical methods and support the development of optical equipment to work with specific spectral bands. • Soil horizon contrast observations are greatly enhanced by reflectance difference analyses. • Spectral indices are accurate on Ratio of Clay prediction (R2 = 0.79; RPD = 2.21). • The present technique is robust to measure ratio of clay, a traditional requisite for soil classification and erosion risk. • The results encourage the creation of field equipments for a quick Ratio of Clay evaluation. [ABSTRACT FROM AUTHOR]