Skip to main content
ARS Home » Southeast Area » Booneville, Arkansas » Dale Bumpers Small Farms Research Center » Research » Publications at this Location » Publication #350996

Research Project: Sustainable Small Farm and Organic Production Systems for Livestock and Agroforestry

Location: Dale Bumpers Small Farms Research Center

Title: Geostatistical mapping of metal elements distribution across conterminous USA

item ADHIKARI, KABINDRA - University Of Arkansas
item Owens, Phillip
item WEST, LARRY - University Of Arkansas
item Ashworth, Amanda
item WILSON, MICHAEL - Natural Resources Conservation Service (NRCS, USDA)
item MILLER, DAVID - University Of Arkansas

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 5/22/2018
Publication Date: 6/1/2019
Citation: Adhikari, K., Owens, P.R., West, L., Ashworth, A.J., Wilson, M.A., Miller, D.M. 2019. Geostatistical mapping of metal elements distribution across conterminous USA. Meeting Abstract. Global Symposium on Soil Pollution. 595-603.

Interpretive Summary:

Technical Abstract: Presence of metals in the soil environment at toxic levels pose human and environmental health concerns. Assessment and proper management of sites contaminated with these metals requires a precise information on the spatial distribution of these metals. This study aims to model and map the spatial distribution of Cd, Cu, Ni, Pb and Zn across the conterminous USA using point observations and geostatistics. About 4,400 surface soil observations of Cd, Cu, Ni, Pb and Zn concentrations (mg kg-1) were compiled from the USDA-NRCS database in this study. Spatial autocorrelation of these elements was quantified with a variogram, and the continuous maps were generated using ordinary kriging. Prediction uncertainty was assessed through prediction standard error map and by cross-validation. Average Cd and Zn in the study area were 0.37 mg kg-1 and 68.6 mg kg-1, respectively; whereas, Cu, Ni, and Pb ranged between 29 to 31 mg kg-1. A stable variogram was found to be the best model to quantify spatial autocorrelation of the elements. Based on cross-validation, Cd had the highest prediction performance and Ni the lowest. In general, the southeastern USA had lower concentrations of most toxic elements compared to the rest of the USA. Cu was higher along the north eastern and west coast; whereas, a toxic concentration of Pb was predicted towards the northeast and a small area in the west north. Results presented in this study provide general trends based on spatially available data. Nevertheless, results from this study give some general information on the spatial behaviours and general distribution pattern of toxic metal elements across the conterminous USA. In the future, mapping with some advanced digital techniques will be adopted in which several soil-environmental variables can be used as predictors of these toxic metals.