Vegetation Recovery Classification from Landsat Imagery in Alberta

Six land cover classes track vegetation recovery from energy exploration disturbances across a 2006-2007 timeframe. The Alberta Energy Regulator created this dataset using Landsat multispectral imagery and land use classification data. It covers a specific pilot area spanning multiple townships in Alberta.

Use Cases

• Classify pixels into one of six recovery types (shrub land, grassland, agricultural areas, coniferous forest, broadleaf forest, mixed forest) using spectral signatures from Landsat imagery.
• Analyze spatial patterns of vegetation recovery following anthropogenic disturbances like oil and gas infrastructure within the defined Play-Based Regulation pilot area.
• Establish a 2006-2007 baseline for change detection studies comparing vegetation loss and recovery over time.
• Validate or train machine learning models for land use and land cover mapping using the provided six-class thematic classification.

Strengths

• Data derived from authoritative Landsat satellite imagery.
• Classification includes six distinct vegetation recovery classes.
• Covers a defined geographic area spanning multiple towns and townships.

Limitations

• Temporal coverage is limited to two years (2006-2007), providing a snapshot rather than a time series.
• Specific sample size, resolution, and accuracy metrics for the classification are not provided.
• Data format includes TIFF files which require geospatial software for full analysis.

Provenance

Source

Alberta Energy Regulator (AER), Government of Alberta.

Collection Method

Produced from Landsat multispectral imagery for 2006 and 2007 and Land Use and Land Cover classification data from 2007.

Time Range

2006-2007

Freshness

null

Geography

Play-Based Regulation pilot area from Twp. 52, Rge. 7, W 5th Mer. to Twp. 70, Rge. 5, W 6th Mer., covering towns including Edson, Fox Creek, Mayerthorpe, Whitecourt, Swan Hills, and Valleyview in Alberta, Canada.