CanLaD: Annual Forest Disturbance Maps for Canada, 1985-2024

Canada Landsat Disturbance (CanLaD) maps forest disturbances across Canada from 1985 to 2024 at a 30-meter resolution. The dataset, produced by Natural Resources Canada, includes annual raster layers classifying disturbance types like wildfire, harvesting, and pest outbreaks, achieving an overall accuracy of 81%. The 2025 update incorporated Sentinel-2 data and manual corrections to reduce false positives.

Use Cases

• Modeling forest carbon flux changes based on annual disturbance maps.
• Analyzing trends in wildfire and pest outbreak severity and location over a 40-year period.
• Monitoring forest harvesting activities and their spatial extent.
• Validating other land cover change products using the dataset's 30m resolution and detailed classification.

Strengths

• Covers a 40-year time series from 1985 to 2024.
• Provides data at a 30-meter spatial resolution.
• Achieves an overall classification accuracy of 81%.
• Includes a 2025 update with corrections using Sentinel-2 data and manual validation.

Limitations

• Pest defoliation classes represent cumulative spectral change, not direct annual aerial survey classifications.
• Models are primarily aimed at conifer pests; detection in deciduous forests may be incidental.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Natural Resources Canada

Collection Method

Based on Landsat time series analyzed with LandTrendr and TempCNN convolutional neural network models.

Time Range

1985 to 2024

Freshness

Last updated 2026-05-14.

Geography

Canada