Liu et al. (2025) GEDI reveals decline in overstorey and increase in understorey canopy cover of protected forests in Central Europe since 2019

Identification

• Journal: Forest Ecology and Management
• Year: 2025
• Date: 2025-09-16
• Authors: Xiao Liu, Vítězslav Moudrý, Bernhard Schuldt, Matthias Forkel
• DOI: 10.1016/j.foreco.2025.123155

Research Groups

• TUD Dresden University of Technology, Institute of Photogrammetry and Remote Sensing, Dresden, Germany
• Czech University of Life Sciences Prague, Faculty of Environmental Sciences, Department of Spatial Sciences, Praha-Suchdol, Czech Republic
• TUD Dresden University of Technology, Institute of Forest Botany and Forest Zoology, Tharandt, Germany

Short Summary

This study utilized GEDI spaceborne lidar data to investigate changes in the vertical forest structure of protected areas in Central Europe since 2019, revealing a widespread decline in overstorey canopy cover coupled with a simultaneous increase in understorey canopy cover in both coniferous and broadleaved forests. The findings highlight GEDI's unique capability to monitor understorey regeneration, providing a more nuanced understanding of forest disturbance and recovery dynamics than traditional forest loss products.

Objective

• Quantify the rate of forest overstorey canopy loss in Central Europe since 2019.
• Quantify the rate of understorey regrowth in Central Europe since 2019.
• Analyse the relationship between forest overstorey loss and understorey regrowth to demonstrate forest recovery across European forests.

Study Configuration

• Spatial Scale: Central Europe, covering protected areas larger than 25 square kilometres in Austria, Czechia, Germany, Hungary, Poland, Slovakia, and Switzerland. GEDI footprints were aggregated monthly per protected area and forest type.
• Temporal Scale: April 2019 to March 2023 for GEDI data acquisition and analysis of canopy cover changes.

Methodology and Data

• Models used:
  • Seasonal and Trend decomposition using Loess (STL) for deseasonalizing monthly GEDI time series.
  • Theil-Sen estimator for quantifying the rate of canopy cover change.
  • Linear regression model for correlation analysis.
  • Random Forest regression model (used in the ancillary Kacic dataset).
• Data sources:
  • Global Ecosystem Dynamic Investigation (GEDI) spaceborne waveform lidar (version 2 Level 2B product) for canopy cover and vertical profile metrics.
  • World Database on Protected Areas (WDPA, January 2024 version) for defining study sites.
  • Copernicus 10-metre resolution forest type product for 2018 (derived from Sentinel-2 satellite) for stratifying by forest type (coniferous, broadleaved).
  • Hansen Global Forest Change 2000-2023 (version 1.11) for comparison of forest loss.
  • Forest Structure Germany dataset (Kacic et al., 2023), derived from Sentinel-1, Sentinel-2, and GEDI, for comparison of total canopy cover changes.

Main Results

• Overstorey canopy cover declined, and understorey canopy cover increased in both coniferous and broadleaved protected forests across Central Europe since 2019.
• In the Harz Mountains, Germany, coniferous overstorey canopy cover decreased by approximately 4.45 % per year, while broadleaved overstorey decreased by approximately 1.57 % per year. Concurrently, understorey canopy cover increased by approximately 1.18 % and 0.59 % per year, respectively.
• For coniferous forests, a significant negative correlation (Pearson r = –0.78, p < 0.01) was found between overstorey canopy cover change and understorey canopy cover change, indicating that a 10 % reduction in overstorey corresponds to an approximate 2.7 % increase in understorey per year.
• No significant linear correlation was observed between overstorey and understorey canopy cover changes in broadleaved forests.
• The observed decline in overstorey canopy cover in coniferous forests was consistent with other remote sensing products (Hansen forest loss: r = –0.89, p < 0.01; Kacic total canopy cover: r = 0.81, p < 0.01). This consistency was weaker for broadleaved forests.
• The increase in understorey canopy cover is primarily driven by forest regeneration and active succession processes, rather than solely increased lidar penetration due to overstorey loss.
• Germany showed the fastest decline in overstorey canopy cover, while Czechia exhibited the fastest increase in understorey canopy cover.

Contributions

• Provides the first large-scale, detailed analysis of vertical forest structure dynamics (overstorey and understorey canopy cover) in Central European protected forests using spaceborne lidar.
• Demonstrates GEDI's unique capability to detect and quantify understorey regrowth, a critical component of forest recovery often missed by passive optical remote sensing.
• Offers a new perspective for monitoring forest disturbance and recovery, complementing existing forest loss products by incorporating the vertical dimension of forest structure.
• Highlights the differential responses of coniferous and broadleaved forests to disturbances, with coniferous forests showing a stronger link between overstorey mortality and understorey regeneration.
• Proposes a methodological framework for integrating GEDI data across mission phases to enable long-term monitoring of forest vertical structure dynamics.

Funding

• China Scholarship Council (Project No. 202006270054)
• EarthBridge project (Building Bridges between Earth Observation and Environmental Sciences, Project No. 101079310) from the European Union Horizon 2020 program.

Citation

@article{Liu2025GEDI,
  author = {Liu, Xiao and Moudrý, Vítězslav and Schuldt, Bernhard and Forkel, Matthias},
  title = {GEDI reveals decline in overstorey and increase in understorey canopy cover of protected forests in Central Europe since 2019},
  journal = {Forest Ecology and Management},
  year = {2025},
  doi = {10.1016/j.foreco.2025.123155},
  url = {https://doi.org/10.1016/j.foreco.2025.123155}
}