Open Collections

Abstract

The historic Cronin Mine (the Site) is located approximately 30 kilometres northeast of Smithers, B.C., and operated intermittently from 1917 to 1974, primarily mining silver, lead and zinc. The Site is divided into two distinct geographical areas: the Upper Mine and the Lower Mine. The Upper Mine comprises the former mine workings and waste rock deposits, and the Lower Mine comprises the former concentrator mill site and tailings deposits. The Site was once accessible via a resource road (RR), however, the RR is no longer passable by vehicle, and the most recent investigations at the Site were completed via helicopter only. The most recent investigation (2024) of the Site focused on the Upper Mine. Situated at elevation(s) ranging from 1,440 to 1,830 metres above sea level, this area is characterised by steep mountainous terrain, and exposed bedrock. A major challenge in investigating the Upper Mine has been distinguishing between elevated metals and metalloids in soil due to historic mining activities (e.g., generation and deposition of waste rock) versus naturally elevated background concentrations from localized mineralization and the physical weathering of the exposed bedrock. Preliminary site investigations to define the spatial distribution(s) of elevated metals and metalloids in soil resulted in the identification of areas that extended beyond the footprint of former mine activities and were too large to physically remediate. Recently, unsupervised machine learning methods combined with multivariate analysis (using R Software) was identified as a tool that could potentially be used to refine these spatial distributions. This paper will describe how the results of the unsupervised machine learning process provided a clear visualization of the spatial distribution of metals and metalloids concentrations in soil and a key tool to advance remedial planning at the Upper Mine, Cronin Mine.