Mapping Soil Properties Research Articles

Increment-averaged kriging for 3-D modelling and mapping soil properties: Combining machine learning and geostatistical methods

Prediction and mapping of soil properties for different soil depths can provide important information for effective land management. Such predictions and maps are often built based on soil data from multiple soil surveys, and the sampled depths will rarely align with depths for which the predictions and maps are required. A recently proposed approach to deal with such datasets, termed increment-averaged kriging (IAK), fits a single model using data from all profiles and all soil depths. The approach considered several potential stumbling blocks: the effect that different widths of the sampled depth intervals has on the variances (sample support); potential non-stationarity in the covariances, depending on soil depth; the need to account for interactions between depth and the effect of covariates. In this work, we present some extensions to the original work that (i) extend the covariance model, (ii) integrate the use of a machine learning algorithm, Cubist, into the framework so that its regression parameters can be fitted while accounting for the correlation and sample support of the data, and (iii) apply a composite likelihood approximation to enable estimation and prediction with large datasets. Code to implement the method in R is also made available so that practitioners can test IAK on their own datasets.

Evaluation of Spatial Variability of Some Soil Properties and Fertility Status Using Nutrient Index and GIS in Bilqas District, Dakahlia Governorate, Egypt

Tracking changes of spatial variability is important to know soil fertility status. A study was conducted to explore the spatial variability of soil properties using GIS, fertility status using nutrient index and changes of spatial variability comparative with a previous study in Bilqas District, Dakahlia Governorate, Egypt. The studied area covers about 684 km2. Accordingly, eleven surface soil samples were collected according to the previous study locations using GPS. The obtained results indicated that, the spatial variability maps of soil properties changed from site to other and from time to other. When comparing the spatial variability of soil properties between the current and previous study, it was observed a decreasing in BD, available NP and CEC in the current study. While, there are a increasing in EC, pH, OM, available K, C/N ratio and ESP. In general, the higher values of clay, exchangeable Ca and Mg, OM, TN, C/N ratio, CEC and ESP were observed in the southern parts of the studied area. In contrary, the lower values of sand, silt, exchangeable K and Na, bulk density, EC and pH were monitor in the southern parts, while the lower values of available NPK were observed in northern parts. Soil fertility status evaluation using nutrient index was low according to TN and available N. While, it was medium according to salt index and exchangeable Ca. Additionally, soil fertility status was high according to soil reaction index, OM, available P, available K, CEC, C/N ratio, exchangeable K, Na and Mg.

Spatial prediction of soil organic carbon stocks in Ghana using legacy data

Soil organic carbon (SOC) is a major driver of the multiple functions of soils in the delivery of ecosystem services. While the spatial prediction and mapping of SOC stocks are of key importance for land managers to understand the synergies between SOC management and soil health, a spatially-explicit map of the distribution of SOC stocks in Ghana is non-existent. Therefore, we quantified the spatial distribution of SOC stocks and associated uncertainties to a target depth of 0–30 cm based on regression-kriging modelling to fill this knowledge gap. The mean error (ME) of the predictions is negligible. The mean absolute error (MAE) shows that the model has prediction errors of about 0.48%. The coefficient of determination (R2) shows that the model explains 34% of the variation in model predictions of SOC stocks. The RMSE is 0.63% of the prediction errors. The predicted SOC stocks show significant variation in their spatial distribution throughout the country. Generally, a trend of decreasing SOC stocks from the southwest to the northeast is clearly recognized. SOC stocks are highest in the Semi-Deciduous agro-ecological zone (43.5 Mg C ha−1) and lowest in the Guinea Savannah agro-ecological zone (0.05 Mg C ha−1). About 5.4 Tg of SOC stocks is stored in the top 0–30 cm of the soils in Ghana. This preliminary work at a spatial resolution of 30 arc-seconds (~1 km) has been accomplished within the framework and guidelines of the Global Soil Partnership (GSP). To our knowledge, this is the first time ever in Ghana that a soil property map has been produced along with its uncertainties. Thus, this study represents a significant first step towards revolutionizing future soil property mapping in Ghana. Even though there remains the need to improve on the quality and spatial resolution, the SOC stocks map presented herein is a satisfactory first step to guide future research on soil organic carbon management at both national and global scales.

Thermal Imaging Techniques in Agricultural Applications

Thermal imaging innovation has transformed into an important technique in all fields as a result of its non-ruinous, non-contact method to choose thermal properties and highlights of any object/zone of interest. It is a procedure of improving the permeability of objects in a dark environment by detecting the objects infrared radiation and shaping an image based on that obtained data. This method will be useful in all fields where temperature difference helps in analyzing the area/object. The potential use of this technique in the agricultural field includes irrigation scheduling, soil properties mapping, plant disease detection, crop yield estimation, evaluating the maturity of the crop, etc. This paper briefly reviews some of the works on thermal imaging for agricultural applications.

Predictive Mapping of Soil Properties for Precision Agriculture Using Geographic Information System (GIS) Based Geostatistics Models

In precision Agriculture, geostatistical methods as a predictive tool have been extensively utilized. The approach estimates soil properties spatial variability and dependency. This study was carried out in Ovia north east Local Government Area of Edo State of Nigeria in order to map soil properties (Sand, Clay, pH, OC, P, N and CEC) and redict their spatial variability. Twenty-nine (29) soil samples were collected randomly from Typic Kandiudults soil type under three different land use, teak forest plantation, shrub, and arable farm. The soil samples were air-dried and passed through a 2 mm sieve before being analyzed for pH(CaCl2), SOC, Sand, Clay, Phosphorus, Nitrogen, and CEC. Generated data were statistically and geostatistically computed to explain the spatial variability of soil properties. The traditional method of soil analysis and interpretation are tedious, time-consuming with escalating budgets thus geostatical approach. Available phosphorus yielded large variability with CV=57.08% followed by clay content with CV=49.03%. Spherical, Gaussian, Hole Effect model, Stable, Exponential and Circular models were fitted for all the soil parameters. The result revealed that soil pH, Sand content, TN and CEC were moderate spatially autocorrelated with nugget/sill value of 0.32, 0.21, 0.49 and 0.30 respectively.  SOC also gave a moderate spatially autocorrelated with nugget/sill value of 0.44. And Clay and Available phosphorus were strong spatially autocorrelated with nugget/sill value of 0.15 and 0.13 respectively. Cross-validation of the output maps using the semivariogram showed that the interpolation models are superior to assuming mean for any unsampled area. The output maps will help soil users within the area to proffer best management technology to improve crop, fiber and water production. 
  

A New Index for Remote Sensing of Soil Organic Carbon Based Solely on Visible Wavelengths

Core Ideas A soil carbon remote sensing index was developed using visible wavelengths. The new index performs as well as or better than existing indices. The new index can be applied to any true color image to map soil organic carbon. Remote sensing is a powerful method for mapping soil properties, such as soil organic carbon (SOC), a key property of soil quality. Spectral remote sensing indices that rely on shortwave‐infrared (SWIR) or near‐infrared (NIR) wavelengths have been developed to quantify spatial patterns in SOC. However, the application of SWIR‐ and NIR‐based indices for quantifying fine‐scale patterns of SOC is limited due to the requirement of high‐resolution multispectral or hyperspectral imagery. Visible wavelengths are measured by virtually all sensors, often at high resolution; hence, development of a visible wavelength–based index can greatly increase the ability to remotely estimate SOC. Here we develop such an index by assessing the relationship between laboratory‐measured SOC and spectral reflectance using 7916 SOC and hyperspectral measurements from the nationwide USDA Rapid Carbon Assessment. Our new SOC index (SOCI) predicts SOC concentrations for the 7916 samples with a RMSE of 1.5%, which is comparable to predictions from the SWIR/NIR ratio (RMSE = 1.3%) and outperforms the predictions of an index based on NIR and red wavelengths (RMSE = 2.8%). We applied the index to a high‐resolution satellite image and tested the ability of the image‐based SOCI to predict measured SOC concentrations for a plowed field in Iowa. Regression models with and without local calibration data accurately predict measured SOC, with RMSE values of ∼0.5%. Given the widespread availability of imagery with spectral data in the visible wavelengths, there is potential to use the SOCI to address a range of soil‐agronomic problems.

Estimating Root Zone Soil Moisture Across the Eastern United States with Passive Microwave Satellite Data and a Simple Hydrologic Model

Root zone soil moisture (RZSM) affects many natural processes and is an important component of environmental modeling, but it is expensive and challenging to monitor for relatively small spatial extents. Satellite datasets offer ample spatial coverage of near-surface soil moisture content at up to a daily time-step, but satellite-derived data products are currently too coarse in spatial resolution to use directly for many environmental applications, such as those for small catchments. This study investigated the use of passive microwave satellite soil moisture data products in a simple hydrologic model to provide root zone soil moisture estimates across a small catchment over a two year time period and the Eastern U.S. (EUS) at a 1 km resolution over a decadal time-scale. The physically based soil moisture analytical relationship (SMAR) was calibrated and tested with the Advanced Microwave Scanning Radiometer (AMSRE), Soil Moisture Ocean Salinity (SMOS), and Soil Moisture Active Passive (SMAP) data products. The SMAR spatial model relies on maps of soil physical properties and was first tested at the Shale Hills experimental catchment in central Pennsylvania. The model met a root mean square error (RMSE) benchmark of 0.06 cm3 cm−3 at 66% of the locations throughout the catchment. Then, the SMAR spatial model was calibrated at up to 68 sites (SCAN and AMERIFLUX network sites) that monitor soil moisture across the EUS region, and maps of SMAR parameters were generated for each satellite data product. The average RMSE for RZSM estimates from each satellite data product is <0.06 cm3 cm−3. Lastly, the 1 km EUS regional RZSM maps were tested with data from the Shale Hills, which was set aside for validating the regional SMAR, and the RMSE between the RZSM predictions and the catchment average is 0.042 cm3 cm−3. This study offers a promising approach for generating long time-series of regional RZSM maps with the same spatial resolution of soil property maps.

Assessing the performance of decision tree and neural network models in mapping soil properties

To build any spatial soil database, a set of environmental data including digital elevation model (DEM) and satellite images beside geomorphic landscape description are essentials. Such a database, integrates field observations and laboratory analyses data with the results obtained from qualitative and quantitative models. So far, various techniques have been developed for soil data processing. The performance of Artificial Neural Network (ANN) and Decision Tree (DT) models was compared to map out some soil attributes in Alborz Province, Iran. Terrain attributes derived from a DEM along with Landsat 8 ETM+, geomorphology map, and the routine laboratory analyses of the studied area were used as input data. The relationships between soil properties (including sand, silt, clay, electrical conductivity, organic carbon, and carbonates) and the environmental variables were assessed using the Pearson Correlation Coefficient and Principle Components Analysis. Slope, elevation, geomforms, carbonate index, stream network, wetness index, and the band’s number 2, 3, 4, and 5 were the most significantly correlated variables. ANN and DT did not show the same accuracy in predicting all parameters. The DT model showed higher performances in estimating sand (R2=0.73), silt (R2=0.70), clay (R2=0.72), organic carbon (R2=0.71), and carbonates (R2=0.70). While the ANN model only showed higher performance in predicting soil electrical conductivity (R2=0.95). The results showed that determination the best model to use, is dependent upon the relation between the considered soil properties with the environmental variables. However, the DT model showed more reasonable results than the ANN model in this study. The results showed that before using a certain model to predict variability of all soil parameters, it would be better to evaluate the efficiency of all possible models for choosing the best fitted model for each property. In other words, most of the developed models are site-specific and may not be applicable to use for predicting other soil properties or other area.

Satellite data integration for soil clay content modelling at a national scale

Soil clay content is a key parameter that influences many other soil properties and processes. The potential of adding new and contemporary satellite data for soil property mapping in France is assessed in this study. The soil property maps used for this analysis were produced within the framework of GlobalSoilMap, which was created to deliver global fine grids of soil properties and associated uncertainties using existing soil information and ancillary data to predict these properties based on digital soil mapping techniques. In this study, we evaluate the added value of Moderate Resolution Imaging Spectroradiometer (MODIS), Project for On-Board Autonomy-Vegetation (PROBA-V), and Sentinel-2 (S2) data for predicting the soil clay content at 90 m resolution for mainland France. The rationale behind adding these data is that satellite images and derived products may enable the biogeochemical characteristics of the earth’s surface to be captured more effectively, which in turn enables more precise predictions of the soil clay content. For this methodology, we i) create composite bare soil mosaics and derive the spectral indices from S2 data acquired during sowing periods from 2016 to 2017, ii) extract the first three principal components of harmonized MODIS and PROBA-V normalized difference vegetation index (NDVI) time series acquired in 2003 and 2016 to represent vegetation changes, and iii) test whether the complementary datasets are able to improve the soil clay information compared to a benchmark value. The soil clay content is obtained by using quantile regression forest (QRF) for each GlobalSoilMap depth interval of 0–5 cm, 5–15 cm, 15–30 cm, 30–60 cm, 60–100 cm, and 100–200 cm along with a 10-fold cross-validation having 10 replicates. The results show that the complementary satellite data improve the clay content estimation on bare soil for the topsoil layers (e.g., 0–30 cm) by increasing the R² and decreasing the bias at averages of 0.05 and 1 g kg−1, respectively. Moreover, the first principal component of the harmonized NDVI data is shown to be the second most important variable for estimating the clay content, as indicated by the QRF models. However, the use of only the satellite data and products as input for the QRF does not yield a satisfactory estimate of the clay content. Finally, this work provides a reference for embedding new remote sensing data in existing national soil inventories and national soil information systems. Further research should incorporate new techniques for considering the spatial–temporal variability of the earth’s surface parameters such as soil moisture and roughness.

Climate and Land-Use Change Effects on Soil Carbon Stocks over 150 Years in Wisconsin, USA

Soil organic carbon is a sink for mitigating increased atmospheric carbon. The international initiative “4 per 1000” aims at implementing practical actions on increasing soil carbon storage in soils under agriculture. This requires a fundamental understanding of the soil carbon changes across the globe. Several studies have suggested that the global soil organic carbon stocks (SOCS) have decreased due to global warming and land cover change, while others reported SOCS may increase under climate change and improved soil management. To better understand how a changing climate, land cover, and agricultural activities influence SOCS across large extents and long periods, the spatial and temporal variations of SOCS were estimated using a modified space-for-time substitution method over a 150-year period in the state of Wisconsin, USA. We used legacy soil datasets and environmental factors collected and estimated at different times across the state (169,639 km2) coupled with a machine-learning algorithm. The legacy soil datasets were collected from 1980 to 2002 from 550 soil profiles and harmonized to 0.30 m depth. The environmental factors consisted of 100-m soil property maps, 1-km annual temperature and precipitation maps, 250-m remote-sensing (i.e., Landsat)-derived yearly land cover maps and a 30-m digital elevation model. The model performance was moderate but can provide insights on understanding the impacts of different factors on SOCS changes across a large spatial and temporal extent. SOCS at the 0–0.30 m decreased at a rate of 0.1 ton ha−1 year−1 between 1850 and 1938 and increased at 0.2 ton ha−1 year−1 between 1980 and 2002. The spatial variation in SOCS at 0–0.30 m was mainly affected by land cover and soil types with the largest SOCS found in forest and wetland and Spodosols. The loss between 1850 and 1980 was most likely due to land cover change while the increase between 1980 and 2002 was due to best soil management practices (e.g., decreased erosion, reduced tillage, crop rotation and use of legume and cover crops).

High-resolution digital soil mapping of multiple soil properties: an alternative to the traditional field survey?

Spatial information on soil particle size distribution and soil organic carbon (SOC) are important for land-use management, environmental models and policy-making. Digital soil mapping (DSM) techniques can quantitatively predict these soil properties using minimal resources. However, DSM has not been adequately evaluated at the farm-scale. The aim of this study was to optimise the DSM framework to produce farm-scale soil maps for 366 ha in the Sandspruit catchment, Western Cape, South Africa. Four feature selection techniques and eight predictive models were evaluated on their ability to predict particle size distribution and SOC. A boosted linear feature selection produced the highest accuracy for all but one soil property. The top-performing predictive models were robust linear models for gravel (ridge regression, RMSE 9.01%, R 2 0.75), sand (support vector machine, RMSE 4.69%, R 2 0.67), clay (quantile regression, RMSE 2.38%, R 2 0.52) and SOC (ridge regression, RMSE 0.19%, R 2 0.41). Random forest was the best predictive model for silt content (RMSE 4.12%, R 2 0.53). This approach appears to be robust for farm-scale soil mapping where the number of observations is often small but high-resolution soil data are required.

Salinity Yield Modeling of the Upper Colorado River Basin Using 30‐m Resolution Soil Maps and Random Forests

AbstractSalinity loading in the Upper Colorado River Basin (UCRB) costs local economies upward of $300 million U.S. dollars annually. Salinity source models have generally included coarse spatial data to represent nonagriculture sources. We developed new predictive soil property and cover maps at 30‐m resolution to improve source representation in salinity modeling. Salinity loading erosion risk indices were also created based on soil properties, remotely sensed bare ground exposure, and topographic factors to examine potential surface soil erosion drivers. These new maps and data from previous SPARROW models were related to recently updated records of salinity at 309 stream gauges in the UCRB using random forest regressions. Resulting salinity yield predictions indicate more diffuse salinity sources, with slightly higher yields in more arid portions of the UCRB, and less overall load coming from irrigated agricultural sources. Model simulations still indicate irrigation to be the major human source of salinity (661,000 Mg or 12%) and also suggest that 75,000 Mg (1.4%) of annual salinity in the UCRB is coming from areas with excessive exposed bare ground in high‐elevation mountain areas. Model inputs allow for field‐scale screening of locations that could be targeted for salinity control projects. Results confirm recent studies indicating limited surface erosional influence on salinity loading in UCRB surface waters, but impacts of monsoonal runoff events are still not fully understood, particularly in drylands. The study highlights the utility of new predictive soil maps and machine learning for environmental modeling.

Assessment of Spatial Variability in Soil Properties of Nagalvadi Micro-Watershed using Geospatial Techniques for Site-Specific Agricultural Input Management in Wardha District of Maharashtra

Soil sampling at many sites is costly and time consuming and hence warrants for predicting the soil properties at un-sampled location from sampled site using spatial dependence characteristics of soil properties through interpolation with reasonable accuracy. Soil samples (0 to 20 cm depth) were collected from a regular grid of 200 by 200 m from Nagalvadi micro-watershed of Wardha district, Maharashtra and analyzed for sand, silt, clay, bulk density, pH, organic carbon, cation exchange capacity, soil moisture retention at -33 kPa and -1500 kPa, available N, P, K and micronutrient cations. Soil thematic maps were generated using semivariogram analysis and ordinary kriging. Spatial variability in soil properties indicated that sand, silt, clay, bulk density, organic carbon, cation exchange capacity, moisture retention at -33 kPa, available N, Fe, and Zn have displayed moderate spatial dependence, whereas, soil pH and available Mn showed strong spatial dependence. Soil organic carbon and soil moisture retention at -33 kPa were spatially correlated for a short range. The kriged maps of soil properties and soil fertility generated are useful for better management decisions.

Application of soil properties, auxiliary parameters, and their combination for prediction of soil classes using decision tree model

Soil classification systems are very useful for a simple and fast summarization of soil properties. These systems indicate the method for data summarization and facilitate connections among researchers, engineers, and other users. One of the practical systems for soil classification is Soil Taxonomy (ST). As determining soil classes for an entire area is expensive, time-consuming, and almost impossible, this research has tried to predict the soil classes in each level of the ST system (up to family level) by using the data of 120 excavated pedons and some auxiliary parameters (such as derivatives of digital elevation model, i.e., DEM) in Shahrekord plain, central Iran. For this reason, the decision tree model was encoded and implemented in the MATLAB software for three conditions: use of soil properties, auxiliary parameters, and its combination. According to the results, soil class prediction error by using soil properties, auxiliary parameters, and its combination was estimated to be 0, 3.33 and 0% for order and suborder levels; 0.83, 15 and 0.83% for great group level; 3.33, 22.5 and 3.33% for subgroup level and 30, 52.5 and 30% for family level, respectively. In addition, the use of kriging maps of soil properties (instead of 120 observational points) decreased the prediction error of the modeling in all levels of the ST system. It seems that the effect of auxiliary parameters (in comparison to soil properties) is not very significant for predicting soil classes in low-relief areas.

Spatial distribution and stoichiometry of soil carbon, nitrogen and phosphorus along an elevation gradient in a wetland in China

AbstractDespite extensive studies on how environmental factors influence plant nutrient distribution and stoichiometry, it remains unclear how elevation affects soil distribution and stoichiometry. Here, patterns of the spatial distribution and stoichiometry of soil carbon (C), nitrogen (N) and phosphorus (P) were studied (with 123 soil samples) along a small‐scale elevation gradient in the wetland of East Dongting Lake, China. Plant nutrient concentrations (C, N and P), soil stoichiometry (C:N, C:P and N:P ratios) and their relations with other soil properties were analysed. In addition, the spatial distributions of plant nutrient concentrations and stoichiometry were evaluated geostatistically with regression kriging, entailing estimation of regression parameters by generalized least squares and ordinary kriging of the residuals. Soil C, N and P concentrations and stoichiometry had moderate to strong spatial dependence, and the ratios of C:P and N:P had similar patterns of distribution to soil organic carbon (SOC) and total nitrogen (TN). The SOC, TN, C:N, C:P and N:P ratios increased with increasing elevation, whereas total phosphorus (TP) showed no marked change. Furthermore, SOC, TN, C:N, C:P and N:P decreased considerably with increasing duration of submergence, soil moisture and pH. Therefore, elevation appeared to influence the distribution of plant nutrients and stoichiometry as a result of changing soil moisture and duration of submergence.Highlights Study showed how elevation influences plant nutrient distribution and stoichiometry. Regression kriging was used to map soil properties. Plant nutrients and stoichiometry decreased with increasing duration of submergence. Elevation affected nutrient distribution because of soil moisture and water regime.

High-resolution three-dimensional mapping of soil organic carbon in China: Effects of SoilGrids products on national modeling

Soil organic carbon (SOC) is a key factor in soil fertility and structure and plays an important role in the global carbon cycle. However, SOC causes a large uncertainty in Earth System Models for predicting future climate change. The GlobalSoilMap (GSM) project aims to provide global digital soil maps of primary functional soil properties at six standard depth intervals (0–5, 5–15, 15–30, 30–60, 60–100, and 100–200 cm) with a grid resolution of 90 × 90 m. Currently, few SOC national products that meet the GSM specifications are available. This study describes the three-dimensional spatial modeling of SOC maps according to GSM specifications. We used 5982 soil profiles collected during the Second National Soil Survey of China, along with 16 environmental covariates related to soil formation. The results were obtained by parallel computing over tiles of 100 × 100 km, and the predictions for the tiles were subsequently merged into a single SOC map for the whole of China per standard GSM depth interval. For each standard GSM depth interval, SOC contents and their uncertainties were predicted and mapped at a spatial resolution of approximately 90 m using bootstrapping. Southwestern and northeastern China had higher SOC contents than the rest of China did, whereas northwestern China had a lower SOC content. The range of the coefficient of determination for the six depth intervals ranged from 0.35 to 0.02, and the mean SOC content was 17.86–8.67 g kg−1. Both these values decreased strongly with increasing soil depth. Cropland SOC content was lower than that of forest and grassland. The results of variable importance show that SoilGrids data were the best predictors for defining the soil-landscape relationship during regression modeling for SOC. These SOC maps can provide a data source for environmental modeling, a benchmark against which to evaluate and monitor SOC dynamics, and a guide for the design of future soil surveys.

Digital mapping of soil ecosystem services in Scotland using neural networks and relationship modelling—Part 1: Mapping of soil classes

AbstractA digital mapping approach was applied to soils in Scotland, producing maps at 100‐m resolution and different levels of classification. This used neural networks to predict fuzzy soil class weightings based upon site descriptors from existing soil survey data. The intention of this work was to produce a set of soil maps for Scotland, which provide greater spatial resolution mapping than currently available, and provide fuzzy data to be used in mapping of ecosystem services using a novel approach explained in a second paper. When selecting the class with the highest weighting for test data points, Producer's Accuracy of 70.7% was achieved in mapping the broadest classification (five classes), while Producer's Accuracy levels of 59.9% and 43.7% were achieved for 11 and 30 classes, respectively. Evaluation of confusion matrices generated from the neural network model tests showed that, particularly for the classification systems with 11 and 30 classes, misclassification errors were more common between similar soil classes than between classes that were very different from one another. This implies that straightforward estimation of classification accuracy that measures “correct” and “incorrect” gives misleading results, and that the fuzzy classification maps are more useful than the crisp classification accuracy values. These fuzzy classification maps are therefore potentially useful for evaluating ecosystem services relating to soil type. We conclude that the mapping approach used here provides a classification accuracy comparable to other approaches and that the outputs can be used for mapping soil properties, processes, functions and ecosystem services.

Modern Trends and Problems of Soil Mapping

The main trends in the development of soil mapping methods are discussed, and the major problems are identified. By the present time, the transition from the paper-based soil maps to digital soil-geographical databases has already been completed. The digital mapping of soils and their properties is now accepted as the main method at all the levels of generalization. The approaches of digital soil mapping, as well as of the traditional one, are based on the ideas of V.V. Dokuchaev about the dependence of soils on soil-forming factors. However, in digital soil mapping, new achievements of mathematical statistics and mathematical modeling are being widely applied. This provides for a greater objectivity and reproducibility of the digital soil maps in comparison with the traditional soil maps. At the same time, all unsolved problems of soil cartography related to the lack of field observation data, scale, soil taxonomy, spatial microheterogeneity, and mapping of individual soil properties are preserved. Partially, these problems can be solved by using remote sensing data. When the soil geographical information is used to assess the quality of soil resources, the interpretation of remote sensing data for mapping purposes seems to be more preferable in comparison with the methods of digital soil mapping.

Relative prediction intervals reveal larger uncertainty in 3D approaches to predictive digital soil mapping of soil properties with legacy data

Fine scale maps of soil properties enable efficient land management and inform earth system models. Recent efforts to create soil property maps from field observations tend to use similar tree-based machine learning interpolation approaches, but often deal with depth of predictions, validation, and uncertainty differently. One of the main differences in approaches is whether to model individual depths of interest separately as ‘2D’ models, or to create models that incorporate depth as a predictor variable creating a ‘3D’ model that can make predictions for all depths. It is unclear how choice of 2D or 3D approach influences model accuracy and uncertainty due to lack of direct comparison and inconsistent presentation of results in past studies. This study compares 2D and 3D methods for mapping soil electrical conductivity (salinity), pH, sum of fine and very fine sands, and organic carbon at 30 m resolution for the upper 432,000 km2 of the Colorado River Watershed of the United States of America. A new, simple, model-agnostic relative prediction interval (RPI) approach to report uncertainty is presented that scales prediction interval width to the 95% interquantile width of the original training sample distribution. The RPI approach enables direct comparison of uncertainty between properties and depths and is easily interpretable by end users. Results indicate that 3D mapping of soil properties with strong variation with depth can result in substantial areas with much higher uncertainty that coincide with unrealistic predictions relative to 2D models, even though 3D models had slightly better global cross-validation scores. Maps and global model summaries of RPI proved helpful in identifying these issues with 3D models. These results suggest that the use of RPI or similar approaches to evaluate models can identify accuracy problems not evident in global validation diagnostics.

POLARIS Soil Properties: 30‐m Probabilistic Maps of Soil Properties Over the Contiguous United States

AbstractSoils play a critical role in the cycling of water, energy, and carbon in the Earth system. Until recently, due primarily to a lack of soil property maps of a sufficiently high‐quality and spatial detail, a minor emphasis has been placed on providing high‐resolution measured soil parameter estimates for land surface models and hydrologic models. This study introduces Probabilistic Remapping of SSURGO (POLARIS) soil properties—a database of 30‐m probabilistic soil property maps over the contiguous United States (CONUS). The mapped variables over CONUS include soil texture, organic matter, pH, saturated hydraulic conductivity, Brooks‐Corey and Van Genuchten water retention curve parameters, bulk density, and saturated water content. POLARIS soil properties was assembled by (1) depth harmonizing and aggregating the pedons in the National Cooperative Soil Survey Soil Characterization Database and the components in Soil Survey Geographic Database into a database of 21,481 different soil series, each soil series having its own vertical profiles of different soil properties, (2) pruning the original POLARIS soil series maps using conventional soil maps to improve soil series prediction accuracy, and (3) merging the assembled soil series databases with the pruned POLARIS soil series maps to construct the soil property maps over CONUS. POLARIS soil properties includes 100‐bin histograms for each layer and variable per grid cell and a series of summary statistics at 30‐, 300‐, and 3,000‐m spatial resolution. Evaluation of POLARIS soil properties using in situ measurements shows an average R2 of 0.41, normalized root‐mean‐square error of 12%, and a normalized mean absolute error of 8.8%.