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.
Citation Information
| Publication Year | 2020 |
|---|---|
| Title | Bayesian modeling of non-stationary, univariate, spatial data for the Earth sciences |
| DOI | 10.3133/tm7C24 |
| Authors | Karl J. Ellefsen, Bradley S. Van Gosen |
| Publication Type | Report |
| Publication Subtype | USGS Numbered Series |
| Series Title | Techniques and Methods |
| Series Number | 7-C24 |
| Index ID | tm7C24 |
| Record Source | USGS Publications Warehouse |
| USGS Organization | Central Mineral and Environmental Resources Science Center; Crustal Geophysics and Geochemistry Science Center |
Related Content
Bayesian modeling of NURE airborne radiometric data for the conterminous United States: predictions and grids
User guide to the bayesian modeling of non-stationary, univariate, spatial data using R language package BMNUS
Related Content
- Data
Bayesian modeling of NURE airborne radiometric data for the conterminous United States: predictions and grids
This data release includes estimates of potassium (K), equivalent uranium (eU), and equivalent thorium (eTh) for the conterminous United States derived from the U.S. Geological Survey's national airborne radiometric data compilation (Duval and others, 2005). Airborne gamma ray spectrometry (AGRS) measures the gamma-rays that are emitted from naturally occurring radioactive isotopes found in rocks - Publications
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 th - Connect