COMPARATIVE ASSESSMENT OF GEOSTATISTICAL TOOLS: A SUDBURY CASE STUDY 1 José F. Saavedra 2 , Graeme Spiers 3 , Paul G. Dunn 4 Abstract: Soil sampling, more than any other medium, reflects the total historical metal accumulation from point sources. In Sudbury, the mining and processing of mineral deposits for over a century has resulted in concerns about the potentially high levels of heavy metals within the soil environment. Sampling, analysis and interpretation of information describing metal levels within the footprint of the Sudbury smelter region is a pre-requisite for future developments in ecological and human risk assessment projects. In this paper a set of tools for incorporating spatial and temporal coordinates of observations into data processing is used to analyze a dataset. The data used is based on a sampling program developed using a randomized stratified sampling plan with modifications to remove, wherever possible, the potential effects of infrastructure. The nested sampling grid, covering an area 200 km by 200 km, was centered on the three smelters; Copper Cliff, Coniston and Falconbridge. The findings from the case study undertaken are that standard spatial interpolation techniques tend to overestimate the metal contents within the soil. The paper also highlights the potential for utilizing simulation and other geostatistical methods for providing the probabilistic distribution of metals within the soil and thereby improving the decision making process. Using the proposed methodology, differences up to 45% can be achieved when considering rehabilitation measures, thus showing the degree of risk the planner could face and how uncertainty could affect decisions. Key Words: Geostatistics, Soil Characterization, Sudbury, Smelter, metal contaminants, spatial sample analyses. Introduction Sudbury has been home to mining, smelting and refining of nickel-copper ores since the late nineteenth century. The early methods of refining included the use roast yards in which layered sulphide ore with locally cut timber was ignited to heat the ore until the sulphide minerals ignited. The resultant nickel and copper concentrates were gathered for further refining (Winterhalder, 1995). The roasting process generated dense plumes of smoke, including sulphur dioxide (Freedman and Hutchinson, 1980). Thus the local forests were denuded by felling for use as fuel and for construction of the railway, and also by the noxious gases emanating from the roast beds. Rapid, severe erosion of the barren soils ensued, resulting in exposure of the bedrock that has, in turn, been subject to intense acid weathering. Estimates suggest that as much as 2.7 x tonnes of SO 5 10 2 were emitted annually, together with many tonnes of heavy metal particulates (Holloway, 1917), at the peak of the ore roast yard era between about 1895 and 1928. In 1928 the use of open roast beds was forbidden by an order from the Ontario Legislature. The open roast beds were supplemented with more efficient smelter facilities with smoke stacks. The three smelters in the Sudbury region were located at Copper Cliff, Coniston and Falconbridge. 1 Paper was presented at Mining and the Environment IV Conference, Sudbury, Ontario, Canada, October 10-27, 2007. 2 jsavedra@mirarco.org . MIRARCO – Mining Innovation, Laurentian University, 935 Ramsey Lake Road, Sudbury, ON, P3E 2C6 3 gspiers@mirarco.org . Ibid. 4 pdunn@mirarco.org . Ibid.