Mapping temperate old-growth forests in Central Europe using ALS and Sentinel-2A multispectral data.

Area of Science:

• Ecology
• Forestry
• Remote Sensing

Background:

• Old-growth forests are vital for biodiversity and carbon sequestration but face threats across Europe.
• Accurate mapping is essential for conservation and sustainable forest management.
• Existing remote sensing methods struggle to capture the complexity of old-growth forest structures.

Purpose of the Study:

• To develop an accurate method for mapping old-growth forests across Europe.
• To assess the effectiveness of combining spectral and structural remote sensing data.
• To support conservation and forest management strategies.

Main Methods:

• Combined Sentinel-2A multispectral imagery (spectral data) with airborne laser scanning (ALS) point clouds (3D structural data).
• Utilized 15 spectral features (bands, vegetation indices, texture) and 4 ALS structural features.
• Employed the random forest algorithm for classification using three distinct datasets.

Main Results:

• The integration of ALS and Sentinel-2A data significantly enhanced old-growth forest classification accuracy.
• Achieved a high F1-score of 92% for the old-growth class.
• Producer's and user's accuracies were 93% and 90%, respectively, highlighting the method's reliability.

Conclusions:

• Forest structure-sensitive features from ALS and Sentinel-2A data are critical for identifying old-growth forests.
• Integrating open-access satellite and airborne data enables broader spatial monitoring of old-growth forests.
• This approach benefits forest managers, stakeholders, and conservationists in preserving these vital ecosystems.