CLIENT: BC Ministry of Forests
PROJECT DESCRIPTION: PRI developed a Drought and Frost Risk Assessment Tool for the BC Ministry of Forests, applying advanced remote sensing techniques to detect and classify drought-stressed areas across various locations in British Columbia. The work was completed in collaboration with Craig DeLong.
CHALLENGE: Accurately identifying drought-stressed and frost-affected trees across large and varied forest landscapes required a scalable and reliable methodology capable of distinguishing individual tree health conditions from aerial imagery — and translating those findings into a broadly applicable classification tool validated against real-world field conditions.
APPROACH: PRI applied object-based image analysis techniques to segment and classify individual trees captured in UAV imagery, characterizing each tree as live or dead based on spectral and structural attributes. This classified dataset was then used to train a machine learning model for detecting drought-stressed areas in Sentinel satellite imagery, extending the tool’s applicability across broader landscapes. Field validation was subsequently carried out to confirm the accuracy of the modeled results and ensure the tool’s reliability for operational use.
OUTCOME: The project delivered a validated, scalable Drought and Frost Risk Assessment Tool for the BC Ministry of Forests, combining high-resolution UAV analysis with satellite-based classification to provide an efficient and accurate means of monitoring forest health and identifying at-risk areas across British Columbia.