ACAN - Open Geospatial Consortium Testbed-16: Machine Learning

This project will explore ML in the context of geospatial standards for a specific theme: wildland fire. Wildland fires represent those that occur in forests, shrublands, and grasslands. In Canada, wildland fires burn on average 2.5 million hectares of forest annually. Large fires of more than 200 hectares in size account for 97% of this annual burned area. Costs for the suppression of such fires alone are significant, ranging from $500 million to $1 billion per year. When considering additional impacts on Canadian communities, citizens, the environment, and economic activity, wildland fires present major challenges to Canada. The work will focus on two components of wildland fire: planning and response. Planning refers to the identification of strategies and approaches to respond to wildland fires before they occur. Response refers to dynamic decision making that occurs throughout the duration of active wildland fires. The response component also encompasses post-event activities (e.g. planning for the long-term aftermath of wildland fire events). For both components, the primary objective of this work will be to determine how geospatial standards, specifically those published by the Open Geospatial Consortium (OGC), can support the application of ML technologies to improve Canada’s ability to manage wildland fire events. It is expected that the outcomes of the project will support multiple uses within the CGDI, NRCan, nationally, and internationally:

Gain immediate insight into the capabilities and potential of ML for supporting wildland fire event planning and response. Identify potential improvements to or the creation of new geospatial standards to better support wildland fire activities. Identify strategies to improve the CGDI’s ability to support planning and response decision making for extreme events, specifically wildland fire. Understand how NRCan and GeoConnections can further improve Canada’s ability to leverage ML technologies in the geospatial context.

As a strategic member of the OGC, it is an inherent interest of NRCan that outcomes of this work will inform future OGC activities that will be of direct benefit to Canada. As such, it is an objective of NRCan that specific deliverables provided through this work have the potential to facilitate additional official OGC activities.