Young Trees Map England: Machine Learning-Based Forest Restock Detection

A machine learning methodology using remote sensing identifies stands where saplings persist in healthy numbers within an eight-year timeframe since planting. The model was funded by DEFRA through the Natural Capital and Ecosystem Assessment programme and developed by the Forestry Commission. Validation recommends using high-confidence thresholds for restock classification, and future integration with the National Forest Inventory aims to improve national woodland indicators.

Use Cases

• Verify government grant compliance for tree planting based on the eight-year monitoring timeframe mentioned in the description.
• Improve national environmental indicators for forest extent and connectivity based on restock status of clearfell polygons.
• Train transferable machine learning models for land-use classification based on multi-year remote sensing data designed to be resilient to climatic variations.

Strengths

• Model validation recommends specific confidence thresholds (over 95% and under 5%) for reliable restock classification.
• Methodology is designed for easy replication and transferability across years by training on multi-year data.
• Project is funded by DEFRA through the Natural Capital and Ecosystem Assessment (NCEA) programme.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect geographic bias inherent to uk_data, focusing solely on England.

Provenance

Source

Forestry Commission

Collection Method

Machine learning methodology using remote sensing data.

Time Range

Uses an eight-year timeframe since planting for monitoring.

Freshness

Last updated 2026-06-29 22:13:46.616712; freshness should be verified.

Geography

England