Geology topics

Karl J. Ellefsen

Publications and source records attributed to Karl J. Ellefsen.

At least 19 recordsLinked to original sources

USGS2026

Effect of mineral deposit data on predictions from the three-part approach to quantitative mineral resource assessment—A study of 16 previous U.S. Geological Survey assessments

The three-part approach to quantitative mineral resource assessment requires information about the properties of undiscovered mineral deposits in an assessment area. These properties are unknown, so the properties of discovered mineral deposits of the same mineral deposit type are used instead. In the three-part approach, these discovered mineral deposits come from around the world, and their properties constitute the pooled data for that mineral deposit type. Alternatively, these discovered mineral deposits could come from the assessment area, and their properties constitute the tract data for that mineral deposit type. Tract data may be more representative of the undiscovered mineral deposits in the assessment area than the pooled data. The goal of this study was to determine whether resource predictions using pooled data are equivalent to resource predictions using tract data. To this end, 16 previous U.S Geological Survey assessments were studied. For each assessment, resources were predicted for one undiscovered mineral deposit in the assessment area. One set of predictions used pooled data, and another used tract data. The two sets of predictions were compared with an equivalence test, using the six assessment statistics that are commonly reported for mineral resource assessments. Practical equivalence is the condition that two corresponding assessment statistics are within a factor of 1.5 of one another. For each of 2 assessments, all 6 assessment statistics were practically equivalent. For both assessments, the assessment statistics from the pooled data, relative to the corresponding assessment statistics from the tract data, ranged from 1.30 times smaller to 1.03 times larger. For each of 14 assessments, 1 or more of the 6 assessment statistics were not practically equivalent. The assessment statistics from the pooled data, relative to the corresponding assessment statistics from the tract data, ranged from 26.6 times smaller to 5.53 times larger. The use of pooled data has been a standard procedure in the three-part approach since at least 1986. The 16 assessments in this study are not a representative sample of those prior assessments that used pooled data. So, it is inappropriate to use the study results to infer whether pooled data affected the resource predictions for those prior assessments.

Scientific Investigations Report↗

USGS2025

Bayesian mapping of regionally grouped, sparse, univariate earth science data

Some earth science data are naturally grouped by region, and it is often desirable to map these data by region. However, if there are only a few samples within each region, then the map should be smoothed in an appropriate way to mitigate the problems that arise from having only a few samples. A smoothing algorithm based on a Bayesian hierarchical model is developed and presented in this report. This algorithm has several features that make it especially suitable for mapping earth science data: it can account for measurements that are censored, it can process multiple datasets with different measurement errors and different censoring thresholds, and it can calculate the uncertainty in any statistic that is mapped. The algorithm is demonstrated by mapping gold concentrations that are measured in streambed sediments in the Taylor Mountains quadrangle in southwestern Alaska.

Alaska↗

USGS2020

Evaluation of the analytical methods used to determine the elemental concentrations found in the stream geochemical dataset compiled for Alaska

A recent U.S. Geological Survey data compilation of stream-sediment geochemistry for Alaska contains decades of analyses collected under numerous Federal and State programs. The compiled data were determined by various analytical methods. Some samples were reanalyzed by a different analytical method than the original, resulting in some elements having concentrations reported by multiple analytical methods. Consideration of the analytical methods used to determine the elemental concentrations is an important step in a mineral prospectivity analysis. We used the compiled data to compare concentrations of barium (Ba), cobalt (Co), copper (Cu), chromium (Cr), nickel (Ni), lead (Pb), and zinc (Zn) determined by different analytical methods to show how simple data comparisons can identify bias and provide a general sense of the comparability of different analytical methods. The elements were selected because they have a range of geochemical properties that may affect the performance of different analytical procedures. Generally, agreement between Ba, Co, Cu, Cr, Ni, Pb, and Zn concentrations is good for most quantitative methods that use a total decomposition of the sample. However, Cr concentrations typically were lower for methods using quantitative-instrumental analysis following a multi-acid dissolution technique that included hydrofluoric acid compared to those using sinter decomposition. Additionally, low- to middle-range concentrations for Co, Cr, Cu, Ni, Pb, and Zn by instrumental neutron activation (NA) and energy-dispersive x-ray spectroscopy (EDX) analyzed by the National Uranium Resource Evaluation (NURE) program have high uncertainty. Concentrations determined by methods that use partial decomposition of the sample generally correspond well to concentrations determined by methods that use a total decomposition technique, except for Ba and Cr. For Ba and Cr, partial decomposition techniques yield lower concentrations than those determined by methods that use a total decomposition technique. Comparison of Ba, Co, Cr, Cu, Ni, Pb, and Zn concentrations determined by semiquantitative visual six-step direct-current arc emission spectrography (ES_SQ) to those determined by quantitative methods using either a total or partial decomposition technique consistently show scatter that exceeds the values expected based on the range represented by the semiquantitative concentration. The data compilation includes a best-value determination that was selected based on the analytical method from the all concentration data for that sample. Ba, Cr, Co, and Zn concentrations determined by NA usually are selected as the best-value determination. However, the NURE-NA method was designed for high throughput and the uncertainty associated with low- and mid-range concentrations is greater than that of the multi-acid method used to reanalyze many samples. Selection of the multi-acid method over the NURE-NA method for Ba, Co, and Zn could be warranted. Additionally, concentrations determined by ES_SQ usually are selected as the best-value determination over all methods that use a partial decomposition of the sample. Substitution of concentrations determined by methods that use a partial decomposition for those of ES_SQ may be warranted for Co, Cu, Ni, Pb, and Zn. Regardless of the selection of the best-value determination, the dataset remains a mixed method dataset and the uncertainty due to differences in analytical methodology must be considered when using the dataset.

Alaska↗

USGS2020

User guide to the bayesian modeling of non-stationary, univariate, spatial data using R language package BMNUS

Bayesian modeling of non-stationary, univariate, spatial data is performed using the R-language package BMNUS. A unique advantage of this package is that it can map the mean, standard deviation, quantiles, and probability of exceeding a specified value. The package includes several R-language classes that prepare the data for the modeling, help select suitable model parameters, and help analyze the results. This user guide describes the BMNUS package and presents step-by-step instructions to model data that accompany the package.

Techniques and Methods↗

USGS2020

Bayesian modeling of non-stationary, univariate, spatial data for the Earth sciences

Some Earth science data, such as geochemical measurements of element concentrations, are non-stationary—the mean and the standard deviation vary spatially. It is important to estimate the spatial variations in both statistics because such information is indicative of geological and other Earth processes. To this end, an estimation method is formulated as a Bayesian hierarchical model. The method represents the spatially varying mean and the spatially varying standard deviation with basis functions; this formulation implicitly accounts for a spatially varying covariance function. A unique advantage of this method is that it can map the mean, the standard deviation, quantiles, and exceedance probabilities. The method is demonstrated by mapping titanium concentrations, which are measured in the coastal plain of the southeastern United States. Various checks demonstrate that the model fits the data and that the estimated statistics are geologically plausible.

Techniques and Methods↗

USGS2019

Geochemical and mineralogical maps, with interpretation, for soils of the conterminous United States

Between 2007 and 2013, the U.S. Geological Survey conducted a low-density (1 site per 1,600 square kilometers, 4,857 sites) geochemical and mineralogical survey of soils in the conterminous United States. The sampling protocol for the national-scale survey included, at each site, a sample from a depth of 0 to 5 centimeters, a composite of the soil A horizon, and a deeper sample from the soil C horizon or, if the top of the C horizon was at a depth greater than 1 meter, a sample from a depth of approximately 80–100 centimeters. The <2-millimeter fraction of each sample was analyzed for a suite of 45 major and trace elements by methods that yield the total or near-total elemental concentration. The major mineralogical components in the samples from the soil A and C horizons were determined by a quantitative X-ray diffraction method using Rietveld refinement. This report presents all the maps and statistical information for each determined element and mineral along with an interpretive section discussing the possible processes that caused the observed national-scale geochemical and mineralogical patterns. Most often, the geochemical and mineralogical patterns reflect the composition of the underlying soil parent material with some modifications caused by leaching of the more mobile elements (for example, calcium and sodium) in the humid areas of the country.

Scientific Investigations Report↗

USGS2019

Effect of size-biased sampling on resource predictions from the three-part method for quantitative mineral resource assessment—A case study of the gold mines in the Timmins-Kirkland Lake area of the Abitibi greenstone belt, Canada:

The three-part method for quantitative mineral resource assessment is used by the U.S. Geological Survey to predict, within a specified assessment area, the number of undiscovered mineral deposits and the quantity of mineral resources in those undiscovered deposits. The effects of size-biased sampling on such predictions are evaluated in a case study that involves gold mines from the Timmins-Kirkland Lake area of the Abitibi greenstone belt, Canada. The gold mines are divided, based upon the time of the assessment, into two groups: existing mines and future mines. The total produced gold for the existing mines are used to predict, with the three-part method, the total produced gold for the future mines. Then the predictions are compared to the known, total produced gold for the future mines. For comparisons using the mean, the predictions are 1.6 to 12 times too high, depending upon the time of the assessment and the probability density function characterizing the total produced gold in the existing mines. For comparisons using the median, the predictions are 1.3 to 10 times too high, depending upon the time of the assessment. The reason for these excessively high predictions is that the three-part method is based on the assumption that the total produced gold from the existing mines is representative of the total produced gold in the future mines; this assumption is inappropriate because of size-biased sampling. There is reason to be concerned that size-biased sampling adversely affected the resource predictions of previous U.S. Geological Survey assessments that were conducted with the three-part method.

Ontario Province, Quebec Province↗

USGS2019

Shear-wave seismic reflection studies of unconsolidated sediments in the near surface

We have successfully applied of SH-wave seismic reflection methods to two different near-surface problems targeting unconsolidated sediments. At the former Fort Ord, where the water table is approximately 30m deep, we imaged aeolian and marine aquifer and aquitard stratigraphy to a depth of approximately 80m. We identified reflections from sand/clay and sand/silt interfaces and we mapped these interfaces along our transects. At an aggregate study site in Indiana, where the water table is at a depth of 1to2m, we imaged stratigraphy in alluvial sand and gravel, and observe a strong reflection from the 32-m-deep bedrock surface. In both cases, we exploited the high resolution potential of SH waves, their insensitivity to water content, and the possibility of reducing Love wave contamination by working along a roadway. We accomplished our results using only sledgehammer sources and simple data processing flows.

California, Indiana↗

USGS2018

Titanium mineral resources in heavy-mineral sands in the Atlantic coastal plain of the southeastern United States

This study examined titanium distribution in the Atlantic Coastal Plain of the southeastern United States; the titanium is found in heavy-mineral sands that include the minerals ilmenite (Fe 2+ TiO 3 ), rutile (TiO 2 ), or leucoxene (an alteration product of ilmenite). Deposits of heavy-mineral sands in ancient and modern coastal plains are a significant feedstock source for the titanium dioxide pigments industry. Currently, two heavy-mineral sands mining and processing operations are active in the southeast United States producing concentrates of ilmenite-leucoxene, rutile, and zircon. The results of this study indicate the potential for similar deposits in many areas of the Atlantic Coastal Plain. This study used the titanium analyses of 3,457 stream sediment samples that were analyzed as part of the U.S. Geological Survey’s National Geochemical Survey program. This data set was analyzed by an integrated spatial modeling technique known as Bayesian hierarchical modeling to map the regional-scale, spatial distribution of titanium concentrations. In particular, clusters of anomalous concentrations of titanium occur: (1) along the Fall Zone, from Virginia to Alabama, where metamorphic and igneous rocks of the Piedmont region contact younger sediments of the Coastal Plain; (2) a paleovalley near the South Carolina and North Carolina border; (3) the upper and middle Atlantic Coastal Plain of North Carolina; (4) the majority of the Atlantic Coastal Plain of Virginia; and (5) barrier islands and stretches of the modern shoreline from South Carolina to northeast Florida. The areas mapped by this study could help mining companies delimit areas for exploration.

Scientific Investigations Report↗

USGS2017

User’s guide for MapMark4—An R package for the probability calculations in three-part mineral resource assessments

MapMark4 is a software package that implements the probability calculations in three-part mineral resource assessments. Functions within the software package are written in the R statistical programming language. These functions, their documentation, and a copy of this user’s guide are bundled together in R’s unit of shareable code, which is called a “package.” This user’s guide includes step-by-step instructions showing how the functions are used to carry out the probability calculations. The calculations are demonstrated using test data, which are included in the package.

Techniques and Methods↗

USGS2017

Probability calculations for three-part mineral resource assessments

Three-part mineral resource assessment is a methodology for predicting, in a specified geographic region, both the number of undiscovered mineral deposits and the amount of mineral resources in those deposits. These predictions are based on probability calculations that are performed with computer software that is newly implemented. Compared to the previous implementation, the new implementation includes new features for the probability calculations themselves and for checks of those calculations. The development of the new implementation lead to a new understanding of the probability calculations, namely the assumptions inherent in the probability calculations. Several assumptions strongly affect the mineral resource predictions, so it is crucial that they are checked during an assessment. The evaluation of the new implementation leads to new findings about the probability calculations,namely findings regarding the precision of the computations,the computation time, and the sensitivity of the calculation results to the input.

Techniques and Methods↗

USGS2017

Rare earth mineral potential in the southeastern U.S. Coastal Plain from integrated geophysical, geochemical, and geological approaches

We combined geophysical, geochemical, mineralogical, and geological data to evaluate the regional presence of rare earth element (REE)−bearing minerals in heavy mineral sand deposits of the southeastern U.S. Coastal Plain. We also analyzed regional differences in these data to determine probable sedimentary provenance. Analyses of heavy mineral separates covering the region show strong correlations between thorium, monazite, and xenotime, suggesting that radiometric equivalent thorium (eTh) can be used as a geophysical proxy for those REE-bearing minerals. Airborne radiometric data collected during the National Uranium Resource Evaluation (NURE) program cover the southeastern United States with line spacing varying from ∼2 to 10 km. These data show eTh highs over Cretaceous and Tertiary Coastal Plain sediments from the Cape Fear arch in North Carolina to eastern Alabama; these highs decrease with distance from the Piedmont. Quaternary sediments along the modern coasts show weaker eTh anomalies, except near coast-parallel ridges from South Carolina to northern Florida. Prominent eTh anomalies are also observed over large riverbeds and their floodplains, even north of the Cape Fear arch where surrounding areas are relatively low. These variations were verified using ground geophysical measurements and sample analyses, indicating that radiometric methods are a useful exploration tool at varying scales. Further analyses of heavy mineral separates showed regional differences, not only in concentrations of monazite, but also of rutile and staurolite, and in magnetic susceptibility. The combined properties suggest the presence of subregions where heavy mineral sediments are primarily sourced from high-grade metamorphic, low-grade metamorphic, or igneous terrains, or where they represent a mixing of these sources. Comparisons between interpreted sources of heavy mineral sands near the Fall Line and igneous and metamorphic Piedmont and Blue Ridge units showed a strong correspondence with rocks closest to the Fall Line and poor correspondence with rocks farther inland. This strongly suggests that the primary source of those heavy minerals, especially monazite, is the rocks that formed the rocky coast that was present during opening of the Atlantic Ocean, which in turn indicates the importance of coastal processes in forming heavy mineral sand concentrations. Furthermore, narrow radiometric eTh and K anomalies are associated with major rivers, indicating limited spatial influence of fluvial processes. Later coastal plain sediment deposition appears to have involved reworking of sediments, providing an “inheritance” of the rocky coast composition that persists for some distance from the Fall Line. However, this inheritance is reduced with distance, and sediments within ∼100 km of the coast in Georgia and Florida exhibit properties indicative of mixing from multiple sources.

GSA Bulletin↗

USGS2017

Feasibility study for the quantitative assessment of mineral resources in asteroids

This study was undertaken to determine if the U.S. Geological Survey’s process for conducting mineral resource assessments on Earth can be applied to asteroids. Successful completion of the assessment, using water and iron resources to test the workflow, has resulted in identification of the minimal adjustments required to conduct full resource assessments beyond Earth. We also identify the types of future studies that would greatly reduce uncertainties in an actual future assessment. Whereas this is a feasibility study and does not include a complete and robust analysis of uncertainty, it is clear that the water and metal resources in near-Earth asteroids are sufficient to support humanity should it become a fully space-faring species.

Open-File Report↗

USGS2016

User’s guide for GcClust—An R package for clustering of regional geochemical data

GcClust is a software package developed by the U.S. Geological Survey for statistical clustering of regional geochemical data, and similar data such as regional mineralogical data. Functions within the software package are written in the R statistical programming language. These functions, their documentation, and a copy of the user’s guide are bundled together in R’s unit of sharable code, which is called a “package.” The user’s guide includes step-by-step instructions showing how the functions are used to cluster data and to evaluate the clustering results. These functions are demonstrated in this report using test data, which are included in the package.

Techniques and Methods↗

USGS2016

Manual hierarchical clustering of regional geochemical data using a Bayesian finite mixture model

Interpretation of regional scale, multivariate geochemical data is aided by a statistical technique called “clustering.” We investigate a particular clustering procedure by applying it to geochemical data collected in the State of Colorado, United States of America. The clustering procedure partitions the field samples for the entire survey area into two clusters. The field samples in each cluster are partitioned again to create two subclusters, and so on. This manual procedure generates a hierarchy of clusters, and the different levels of the hierarchy show geochemical and geological processes occurring at different spatial scales. Although there are many different clustering methods, we use Bayesian finite mixture modeling with two probability distributions, which yields two clusters. The model parameters are estimated with Hamiltonian Monte Carlo sampling of the posterior probability density function, which usually has multiple modes. Each mode has its own set of model parameters; each set is checked to ensure that it is consistent both with the data and with independent geologic knowledge. The set of model parameters that is most consistent with the independent geologic knowledge is selected for detailed interpretation and partitioning of the field samples.

Colorado↗

USGS2016

Coastal deposits of heavy mineral sands; Global significance and US resources

Ancient and modern coastal deposits of heavy mineral sands (HMS) are the principal source of several heavy industrial minerals, with mining and processing operations on every continent except Antarctica. For example, HMS deposits are the main source of titanium feedstock for the titanium dioxide (TiO 2 ) pigments industry, obtained from the minerals ilmenite (Fe 2 +TiO 3 ), rutile (TiO 2 ) and leucoxene (an alteration product of ilmenite). HMS deposits are also the principal source of zircon (ZrSiO 4 ), from which zirconium dioxide (ZrO 2 ) is obtained for uses mostly in refractory products. Sometimes monazite [(Ce,La,Nd,Th)PO 4 ] is recovered as a byproduct mineral, sought for its rare earth elements and thorium (Ault and others, 2016; Sengupta and Van Gosen, 2016; Van Gosen and Tulsidas, 2016).

Mining Engineering↗

USGS2015

First steps of integrated spatial modeling of titanium, zirconium, and rare earth element resources within the Coastal Plain sediments of the southeastern United States

The Coastal Plain of the southeastern United States has extensive, unconsolidated sedimentary deposits that are enriched in heavy minerals containing titanium, zirconium, and rare earth element resources. Areas favorable for exploration and development of these resources are being identified by geochemical data, which are supplemented with geological, geophysical, hydrological, and geographical data. The first steps of this analysis have been completed. The concentrations of lanthanum, yttrium, and titanium tend to decrease as distance from the Piedmont (which is the likely source of these resources) increases and are moderately correlated with airborne measurements of equivalent thorium concentration. The concentrations of lanthanum, yttrium, and titanium are relatively high in those watersheds that adjoin the Piedmont, south of the Cape Fear Arch. Although this relation suggests that the concentrations are related to the watersheds, it may be simply an independent regional trend. The concentration of zirconium is unrelated to the distance from the Piedmont, the equivalent thorium concentration, and the watershed. These findings establish a foundation for more sophisticated analyses using integrated spatial modeling.

Open-File Report↗

USGS2014

A modified procedure for mixture-model clustering of regional geochemical data

A modified procedure is proposed for mixture-model clustering of regional-scale geochemical data. The key modification is the robust principal component transformation of the isometric log-ratio transforms of the element concentrations. This principal component transformation and the associated dimension reduction are applied before the data are clustered. The principal advantage of this modification is that it significantly improves the stability of the clustering. The principal disadvantage is that it requires subjective selection of the number of clusters and the number of principal components. To evaluate the efficacy of this modified procedure, it is applied to soil geochemical data that comprise 959 samples from the state of Colorado (USA) for which the concentrations of 44 elements are measured. The distributions of element concentrations that are derived from the mixture model and from the field samples are similar, indicating that the mixture model is a suitable representation of the transformed geochemical data. Each cluster and the associated distributions of the element concentrations are related to specific geologic and anthropogenic features. In this way, mixture model clustering facilitates interpretation of the regional geochemical data.

Colorado↗