The legacy of an estimated twenty-seven thousand orphaned or abandoned mines pose environmental, health, safety and economic problems to communities, the mining industry and governments across Canada. Information gathering for these sites is necessary to enable sound decision-making, cost-efficient planning and sustainable rehabilitation. The SDKI Sustainable Management and Rehabilitation of Mine Sites for Decision Support project is working collaboratively with federal departments, provincial governments and industry to develop new techniques for information collection and integration to support mine reclamation and policy decisions surrounding mine rehabilitation.

Figure 1. Photograph of "kill zone" area around KamKotia mine. Acid mine drainage produced toxic soils that were unable to support the existing forest cover.

In support of the Ontario Ministry of Mines and Northern Development (OMND)'s Abandoned Mine Rehabilitation Program, SDKI scientists are providing a case study using these new techniques at the KamKotia mine in northern Ontario. Copper and zinc were mined at KamKotia from 1942 - 1972, before it was abandoned and left under government stewardship. During this period, three million tons of tailings were discharged in the area around the mine, before an impoundment was built in 1967. This produced a "kill zone" of approximately 170 hectares (Fig. 1). In 2001, the OMND began a multi-year project to rehabilitate the site. To date, a water-treatment plant and a new tailings impoundment have been built, and the exposed tailings areas are being dredged and deposited in the new impoundment.

Figure 2. Map of ground cover features over a section of the tailings at the KamKotia mine. This map will serve as a baseline to determine the success of rehabilitation efforts at the site.

A baseline map of the site was developed using information extraction techniques for remote sensing data, developed in collaboration with the Georesearch Centre (GFZ) in Germany (Fig. 2). The map was derived from airborne "hyperspectral" remote sensing data collected in 2001, with a five metre spatial resolution. Hyperspectral sensors collect reflected radiation from the earth's surface in a large number of narrow spectral bands, for example, in the visible and infrared wavelengths. Using automated techniques, this data can be processed to uniquely identify land cover classes on the surface at "sub-pixel" accuracy. The user can then establish thresholds to define applicable management zones. For this example, the study area was first classified into multiple maps depicting the percent cover of vegetation, tailings, and dead vegetation, and these maps were thresholded and combined to produce the final zonation shown in Figure 2.

This map will serve as a basis for comparison with maps produced by SDKI in subsequent years, and will be used to evaluate the success of reclamation efforts. Sites such as KamKotia are significant financial and environmental liabilities to both federal and provincial governments. The use of these techniques to extract information will provide a significant means of ensuring costly rehabilitation actions are effectively meeting established reclamation goals.