Vegetation Recovery Classification From Landsat Imagery in Cold Lake Oil Sands

Six land cover classes—shrub land, grassland, agricultural areas, coniferous forest, broadleaf forest, and mixed forest—are used to classify vegetation recovery from anthropogenic disturbances in Alberta's Cold Lake oil sands region. The dataset was produced by the Government of Alberta using Landsat multispectral imagery from 2008 and 2009. It supports the monitoring objectives of the Lower Athabasca Regional Plan.

Use Cases

• Classify pixels into one of six land cover types (e.g., shrub land, coniferous forest) to map vegetation recovery patterns.
• Analyze changes in vegetation classes between 2008 and 2009 imagery to quantify recovery rates from disturbances.
• Train a semantic segmentation model on Landsat TIFF imagery to predict land cover classes for new areas.
• Assess the spatial distribution of anthropogenic disturbance recovery by analyzing classified forest and grassland areas.

Strengths

• Classification into six distinct land cover classes provides structured analysis of vegetation types.
• Derived from Landsat multispectral imagery, a standardized and well-calibrated satellite data source.

Limitations

• Unknown sample size and spatial resolution limit statistical and granular analysis.
• Data is temporally limited to a two-year period (2008-2009), preventing long-term trend analysis.
• Potential class imbalance or label noise inherent in remote sensing classification is not documented.

Provenance

Source

Government of Alberta

Collection Method

Derived from Landsat multispectral imagery for 2008 and 2009 and Land Use and Land Cover classification data from 2009.

Time Range

2008 to 2009

Freshness

Data reflects conditions from 2008-2009, though metadata was last updated in 2026.

Geography

Cold Lake oil sands area, Alberta, Canada (Township 56 to 69, Range 1 to 11, west of the 4th Meridian).