Land Disturbances and Vegetation Regeneration of Alberta, Derived from 1985 to 2025 Landsat and Sentinel-2 Multispectral Data (image data, GeoTIFF format)

This digital data package contains gridded datasets derived from selected multispectral satellite imagery captured from 1985 to 2025. The datasets describe land disturbance, vegetation regeneration, vegetation regeneration productivity, and possible undesired vegetation conditions across Alberta.

Disturbances were detected through time-series comparison of selected satellite datasets and classified based on the timing and magnitude of spectral change on the landscape. They can be caused by a range of factors, including forest harvesting, oil and gas activities, other infrastructure development, wildfire, and insect defoliation. Disturbances associated with agricultural activities were excluded.

In addition to mapping land disturbance, the dataset identifies vegetation regeneration, which reflects recovery processes that occur naturally or through anthropogenic intervention. Regeneration is typically observed in areas undergoing reclamation activities, forestry cutblocks, and recovery following wildfire.

Four thematic data products were generated from two satellite data sources (Landsat and Sentinel-2), resulting in a total of eight products. Landsat-based products were created at 30 m resolution for 1985–2025 at five-year intervals, whereas Sentinel-2-based products were created at 10 m resolution annually from 2017–2025:

1. magnitude of cumulative land disturbance (Landsat_AB_cumul_dist_mag_1985_2025_class_10tm.tif, Sentinel2_AB_cumul_dist_mag_2017_2025_class_10tm.tif),
2. year of highest disturbance (Landsat_AB_max_dist_yr_1985_2025_class_10tm.tif, Sentinel2_AB_max_dist_yr_2017_2025_class_10tm.tif),
3. magnitude of cumulative vegetation regeneration (Landsat_AB_veg_regen_mag_1985_2025_class_10tm.tif, Sentinel2_AB_veg_regen_mag_2017_2025_class_10tm.tif), and
4. year of highest regeneration (Landsat_AB_max_regen_yr_1985_2025_class_10tm.tif, Sentinel2_AB_max_regen_yr_2017_2025_class_10tm.tif).

The Landsat and Sentinel-2 vegetation regeneration magnitude products are further classified into vegetation regeneration productivity classes using cumulative regeneration magnitude, 2025 seasonal productivity, and the undesired vegetation indicator derived from Sentinel-2 imagery. The vegetation regeneration productivity classes are:

0. No regeneration
1. Undesired vegetation (noxious weed indicator)
2. Very low regeneration productivity
3. Low regeneration productivity
4. Moderate regeneration productivity
5. High regeneration productivity
6. Very high regeneration productivity

The undesired vegetation class was developed using Sentinel-2 spectral unmixing to identify areas with spectral characteristics consistent with possible undesired vegetation, including noxious weed conditions within the boreal forest. This class is intended as a screening indicator to identify areas that may warrant closer review and should not be interpreted as a field confirmed weed inventory.

The Landsat and Sentinel-2 land disturbance magnitude products are classified into ordinal disturbance magnitude classes ranging from no disturbance to minimal, very low, low, moderate, high, very high, very severe, and extreme disturbance magnitude. These classes represent the relative strength of cumulative spectral change associated with disturbance through time.