David B. Smith, U.S. Geological Survey (United States)
Dennis R. Helsel, U.S. Geological Survey (United States)
L. Graham Closs, Colorado School of Mines (United States)
James E. Kilburn, U.S. Geological Survey (United States)
Steven M. Smith, U.S. Geological Survey (United States)
John D. Horton, U.S. Geological Survey (United States)
In 2005, the U.S. Geological Survey and the Colorado School of Mines conducted a geochemical survey of roadside soils in the Denver, Colorado (USA) metropolitan area. The top 12 cm of soil was collected at 497 sites uniformly distributed throughout 1165 km2 of the urban area. Two size fractions, <2 mm and <250 m, of each sample were analyzed for near-total concentrations of 44 elements using, primarily, ICP-MS and ICP-AES following a four-acid extraction. The 2005 data set and geochemical maps were then compared to similar data and maps generated from 439 samples of <250-m roadside soils collected at a similar depth in 1972 by Skyline Laboratories of Wheat Ridge, Colorado. Both the 2005 and 1972 samples were analyzed during 2005 by the same analytical protocols.
These two data sets provide a unique opportunity to map the abundance and spatial distribution of potentially harmful elements in Denver soils at two points in time and to observe possible changes in soil geochemistry that may have occurred during a 33-year time span. Such urban geochemical maps are important for city authorities in understanding the nature and extent of urban contamination in light of such activities as the development of brownfield sites. Several elements show elevated concentrations in the Denver area in both data sets compared to rural background concentrations determined in previous geochemical surveys. These elements include Pb, Zn, As, Hg, Cd, Cu, and Sb. Potential sources for these elements include 1) atmospheric contamination from industrial plants and smelters, 2) automobile exhaust, 3) debris from automobile engines, catalytic converters, tires, and brake pads, and 4) contamination of alluvial soils from mineral deposits and mining activity west of the Denver area.
Quantifying change by comparison of the 2005 and 1972 data sets is not straightforward. Soil exhibits heterogeneity in both the composition of the constituent particles and the spatial distribution of these particles. As a result, all soil geochemical surveys are subject to sampling errors. This is especially true of a survey involving roadside soils as collected in this study. Despite these inherent errors, some observed changes may be significant. For example, Pb exhibited the greatest decrease in concentration from 1972 to 2005, most likely a result of the phasing out of leaded gasoline beginning in the 1970s.