Geostatistical analysis and integration of soil chemistry data with remote sensing information in the Sudbury area, Ontario.
Date
2022-03-28
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The presence of anomalous concentrations of metals within the soil profile can strongly affect its
biological availability to plants, causing potential toxicity when exceeding threshold concentrations,
and favoring numerous chemical exchanges. These interactions further facilitate metal dispersion in
the hydrogeological and ecological systems, in response to weathering and erosion. Studies of the
geospatial distribution of trace metal contaminants in Sudbury soils is thus important to unravel the
dominant processes controlling dispersion patterns, contributing to sustainability of mining practice.
A kriging geostatistical approach was applied to geochemical data obtained from the Sudbury Soil
Survey to map multiscale geographic, enrichment trends in metal concentrations. Ordinary kriging
prediction maps were developed to re-evaluate the multiscale spatial distribution of the chemicals of
concern. Results show an anomalous distribution of metals centered on historical smelters, forming
dominant northeast and southwest enrichment trends. The existence of these trends was validated by
implementing a geostatistical Gaussian conditional simulation method, which reproduced the same
spatial variability observed in the ordinary kriging maps and efficiently replicated the observed
trends. The correlation analysis of the trends with remote sensing data, suggests that prevailing wind
directions are likely one of the dominant driving forces controlling the trends. Integrating these
results with satellite data showed improved vegetation regrowth patterns consistent with the
geochemical northeast-southwest trend providing further, independent validation of the kriging
results. Re-evaluation of the regional, geospatial distribution of the measured trace element
concentrations will assist the monitoring and improved understanding of soil contamination trends and
their impact on vegetation and other aspects of the biosphere in the Greater Sudbury area.
Description
Keywords
Sudbury soil study, Kriging interpolation, Gaussian conditional simulation, remote sensing, normalized difference vegetation index