The Porcupine Gold Camp in Eastern Canada is one of the most prolific gold mining districts in North America, with past production exceeding 60 million ounces of gold. Mining and exploration have been going on in this district since gold was first discovered near Timmins in 1907.
The Abitibi greenstone belt, which contains the Porcupine Camp, is the largest greenstone belt in the world. The major gold camps are spatially associated with steeply dipping shear zones, such as the Destor-Porcupine Fault. It has a high proportion of supracrustal rocks, a generally low metamorphic grade, and contains a wide variety of mineral deposits.
The Porcupine study covers roughly 8,000 square kilometers, which is a much smaller area than the other two case studies. At the time of the study the Porcupine Joint Venture (PJV) was a joint venture between Placer Dome and Kinross Gold Corporation, who were actively exploring for more gold. The available exploration data sets included geology and structure, high quality airborne magnetics and gravity, SRTM data, and a Geotem survey. In addition, the footprints of all the known gold deposits in the Porcupine camp were carefully established by PJV staff.
Data Mining Results
The Porcupine Gold Camp is well-suited to this type of objective analysis. The accompanying figures show a closeup of the results obtained in one part of the camp. The upper figure shows the geology, and the lower figure shows our corresponding target map. The green crosses mark known deposits, and the red crosses mark targets having a greater than one-in-four chance of containing a gold deposit, according to the study. The blue circles mark prior drill holes which have been analyzed for lithogeochemistry and gold concentration.
A valuable test was conducted by Dr. Cliff Saunders, an independent consultant working for Kinross Gold Corporation. He selected a number of sites from the data mining target map, at some of which the process predicted gold, and at others where it predicted there should be no gold. With the help of the PJV geological staff, he examined the drill logs and other geological information from these locations.
At four sites where the process predicted gold, two were found to contain economic gold deposits. The other two did not contain economic gold, but the geologists felt both were good places to look. At four sites where the process predicted no gold, one contained a small gold deposit. At each of the other three sites, extensive drilling had encountered no gold, as predicted.
Apart from the one miss of a small economic deposit, and the couple of sites that were considered prospective by the process but were deemed unfavourable by the geologists, the data mining process gave very good results. In particular, it achieved better than the predicted one-in-four success rate for the occurrence of economic gold.