Geology topics

Colby W. Brungard

Publications and source records attributed to Colby W. Brungard.

5 recordsLinked to original sources

USGS2019

POLARIS properties: 30-meter probabilistic maps of soil properties over the contiguous United States

Soils 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%.

Water Resources Research↗

USGS2018

Approaches for improving field soil identification

Use of soil survey information by non-soil-scientists is often limited by their inability to select the correct soil map unit component (COMP). Here, we developed two approaches that can be deployed to smartphones for non-soil-scientists to identify COMP from the location alone or location together with easily observed field data (i.e., slope, depth to the restrictive layer, and soil texture by depth). In addition, we also compared the two newly developed approaches with a traditional approach identifying COMP based on the dominant COMP (DC-based approach). All three approaches were tested with the Rapid Assessment of US Soil Carbon database and the combined USDA- NRCS Soil Survey Geographic database and the USDA-NRCS State Soil Geographic Database. The results indicated that the observation-based approach performed significantly better than the other two approaches, suggesting that a small set of easy-to-measure site-specific observations could significantly improve COMP identification. The location- and DC-based approaches had similar low performance overall. However, the location-based approach slightly improved identifications over the DC-based approach for cases where (i) there were multiple possible components within the soil map unit and (ii) the components were located in close proximity to a boundary of a different soil map unit polygon. The benefit of using the location-based approach may be greater in specific soil survey areas where topography was the major factor leading to the creation of the map unit legend.

Soil Science Society of America Journal↗

USGS2018

Soil property and class maps of the conterminous United States at 100-meter spatial resolution

With growing concern for the depletion of soil resources, conventional soil maps need to be updated and provided at finer and finer resolutions to be able to support spatially explicit human–landscape models. Three US soil point datasets—the National Cooperative Soil Survey Characterization Database, the National Soil Information System, and the Rapid Carbon Assessment dataset—were combined with a stack of over 200 environmental datasets and gSSURGO polygon maps to generate complete coverage gridded predictions at 100-m spatial resolution of six soil properties (percentage of organic C, total N, bulk density, pH, and percentage of sand and clay) and two US soil taxonomic classes (291 great groups [GGs] and 78 modified particle size classes [mPSCs]) for the conterminous United States. Models were built using parallelized random forest and gradient boosting algorithms as implemented in the ranger and xgboost packages for R. Soil property predictions were generated at seven standard soil depths (0, 5, 15, 30, 60, 100, and 200 cm). Prediction probability maps for US soil taxonomic classifications were also generated. Cross validation results indicated an out-of-bag classification accuracy of 60% for GGs and 66% for mPSCs; for soil properties, RMSE for leave-location-out cross-validation was 0.74 ( R 2 = 0.68), 17.8 wt% ( R 2 = 0.57), 12 wt% ( R 2 = 0.46), 3.63 wt% ( R 2 = 0.41), 0.2 g cm −3 ( R 2 = 0.42), and 0.27 wt% ( R 2 = 0.39) for pH, percent sand and clay, weight percentage of organic C, bulk density, and weight percentage of total N, respectively. Nine independent validation datasets were used to assess prediction accuracies for soil class models, and results ranged between 24 and 58% and between 24 and 93% for GG and mPSC prediction accuracies, respectively. Although mapping accuracies were variable and likely lower than gSSURGO in some areas, this modeling approach can enable easier integration of soil information with spatially explicit models compared with multicomponent map units.

Soil Science Society of America Journal↗

USGS2016

POLARIS: A 30-meter probabilistic soil series map of the contiguous United States

A new complete map of soil series probabilities has been produced for the contiguous United States at a 30 m spatial resolution. This innovative database, named POLARIS, is constructed using available high-resolution geospatial environmental data and a state-of-the-art machine learning algorithm (DSMART-HPC) to remap the Soil Survey Geographic (SSURGO) database. This 9 billion grid cell database is possible using available high performance computing resources. POLARIS provides a spatially continuous, internally consistent, quantitative prediction of soil series. It offers potential solutions to the primary weaknesses in SSURGO: 1) unmapped areas are gap-filled using survey data from the surrounding regions, 2) the artificial discontinuities at political boundaries are removed, and 3) the use of high resolution environmental covariate data leads to a spatial disaggregation of the coarse polygons. The geospatial environmental covariates that have the largest role in assembling POLARIS over the contiguous United States (CONUS) are fine-scale (30 m) elevation data and coarse-scale (~ 2 km) estimates of the geographic distribution of uranium, thorium, and potassium. A preliminary validation of POLARIS using the NRCS National Soil Information System (NASIS) database shows variable performance over CONUS. In general, the best performance is obtained at grid cells where DSMART-HPC is most able to reduce the chance of misclassification. The important role of environmental covariates in limiting prediction uncertainty suggests including additional covariates is pivotal to improving POLARIS' accuracy. This database has the potential to improve the modeling of biogeochemical, water, and energy cycles in environmental models; enhance availability of data for precision agriculture; and assist hydrologic monitoring and forecasting to ensure food and water security.

Geoderma↗

USGS2015

Machine learning for predicting soil classes in three semi-arid landscapes

Mapping the spatial distribution of soil taxonomic classes is important for informing soil use and management decisions. Digital soil mapping (DSM) can quantitatively predict the spatial distribution of soil taxonomic classes. Key components of DSM are the method and the set of environmental covariates used to predict soil classes. Machine learning is a general term for a broad set of statistical modeling techniques. Many different machine learning models have been applied in the literature and there are different approaches for selecting covariates for DSM. However, there is little guidance as to which, if any, machine learning model and covariate set might be optimal for predicting soil classes across different landscapes. Our objective was to compare multiple machine learning models and covariate sets for predicting soil taxonomic classes at three geographically distinct areas in the semi-arid western United States of America (southern New Mexico, southwestern Utah, and northeastern Wyoming). All three areas were the focus of digital soil mapping studies. Sampling sites at each study area were selected using conditioned Latin hypercube sampling (cLHS). We compared models that had been used in other DSM studies, including clustering algorithms, discriminant analysis, multinomial logistic regression, neural networks, tree based methods, and support vector machine classifiers. Tested machine learning models were divided into three groups based on model complexity: simple, moderate, and complex. We also compared environmental covariates derived from digital elevation models and Landsat imagery that were divided into three different sets: 1) covariates selected a priori by soil scientists familiar with each area and used as input into cLHS, 2) the covariates in set 1 plus 113 additional covariates, and 3) covariates selected using recursive feature elimination. Overall, complex models were consistently more accurate than simple or moderately complex models. Random forests (RF) using covariates selected via recursive feature elimination was consistently the most accurate, or was among the most accurate, classifiers between study areas and between covariate sets within each study area. We recommend that for soil taxonomic class prediction, complex models and covariates selected by recursive feature elimination be used. Overall classification accuracy in each study area was largely dependent upon the number of soil taxonomic classes and the frequency distribution of pedon observations between taxonomic classes. Individual subgroup class accuracy was generally dependent upon the number of soil pedon observations in each taxonomic class. The number of soil classes is related to the inherent variability of a given area. The imbalance of soil pedon observations between classes is likely related to cLHS. Imbalanced frequency distributions of soil pedon observations between classes must be addressed to improve model accuracy. Solutions include increasing the number of soil pedon observations in classes with few observations or decreasing the number of classes. Spatial predictions using the most accurate models generally agree with expected soil–landscape relationships. Spatial prediction uncertainty was lowest in areas of relatively low relief for each study area.

New Mexico, Utah, Wyoming↗