June 13, 2018
Machine learning pinpoints mineral deposits in northwest British Columbia

 

Vancouver, BC - June 13, 2018 - A new report published by Geoscience BC today matches machine learning and interpretive techniques for stream sample geochemistry to locate mineral deposits in northwest British Columbia.

The Integrated Assessment of Regional Stream Sediment Geochemistry for Metallic Deposits in Northern Stikine Terrane project used techniques including 'random forest' and 'typicality'/'allocation' supervised machine learning to analyse almost 15,000 geochemical samples from across northwest British Columbia. This identified 'background' geochemistry so that any anomalies indicating mineral deposits are highlighted.

Lead researcher and CSA Global Canada Geosciences Ltd. Principal Consultant Geochemist Dennis Arne said: "Using a combination of classic and modern geochemical data processing techniques allowed us to focus in on the trace elements left by specific mineral deposits by removing the influence of the uninteresting rocks within a stream's catchment area and the impact of metal scavenging and dilution."

Sand, clay and silts collected from a single point in a stream are a mix of materials from further upstream, including tributaries, and so are often used to identify potential mineral deposits. This research amplifies the most useful results and focuses on the elements that point to certain deposit types by applying a mathematical 'filter' to take a fresh look at old data. The research team also used machine learning techniques to distinguish between different mineral deposit types and recommend areas for follow-up investigation.

Commenting on the new report, Geoscience BC Vice President of Minerals and Mining Bruce Madu said: "These complex statistical analyses untangle the data and remove the confusing influences of the rocks in the area. Select information from this project has been added to the Geoscience BC Earth Science Viewer online application to make it easy to access. This helps the resource sector, communities, First Nations and government to make decisions about mineral exploration and land management."

This project used data collected and analysed as part of seven previous Geoscience BC, British Columbia Geological Survey and Geological Survey of Canada projects in a large area that includes communities such as Burns Lake, Dease Lake, Fort St James, Fraser Lake, Houston, Iskut, Kitimat, Smithers, Takla Landing and Terrace.

Geoscience BC has supported similar projects to analyse the catchment areas of stream sediment samples in the past. These include:

Accessing Information

To view the reports and maps, visit the project page or view the information on our Earth Science Viewer.

View 2016-028 project page View 2016-028 in Earth Science Viewer

About Geoscience BC

Geoscience BC is an independent, non-profit organization that generates earth science information in collaboration with First Nations, local communities, governments, academia and the resource sector. Our independent earth science enables informed resource management decisions. You can view a copy of the Geoscience BC Strategic Plan 2018-2022 here. Geoscience BC gratefully acknowledges the financial support of the Province of British Columbia.

For additional information, please contact:

Richard Truman
Geoscience BC
604-662-4147



Follow us on social media
Facebook
Twitter
LinkedIn

 
 

You can view the Next News item: Mon Jun 18, 2018, Geoscience BC at Resources for Generations (RFG)

You can view the Previous News item: Thu May 24, 2018, Geoscience BC is hiring!

You can return to the main News page, or press the Back button on your browser.



Show printable version of 'Machine learning pinpoints mineral deposits in nor...' in a New Window