Liang et al. (2026) Vegetation canopy height retrieval in complex mountainous regions based on data calibration and CNN model

Identification

• Journal: International Journal of Applied Earth Observation and Geoinformation
• Year: 2026
• Date: 2026-09-18
• Authors: Qiang Bie, Hongwei Zhang, Wenyu Yao
• DOI: 10.1016/j.jag.2026.105598

Research Groups

• Faculty of Geomatics, Lanzhou Jiaotong University
• Gansu Provincial Engineering Laboratory for National Geographic State Monitoring
• Key Laboratory of Urban Land Resources Monitoring and Simulation, Ministry of Natural Resources

Short Summary

This study proposes a regional calibration approach for GEDI-derived height labels and integrates the calibrated labels with multi-source remote sensing data to retrieve canopy height at 10 m resolution. The optimized model achieves a mean absolute error (MAE) of 2.26 m against field measurements, showing strong consistency with airborne LiDAR validation.

Objective

• To develop an accurate method for retrieving vegetation canopy height in complex mountainous regions using GEDI data calibration and multi-source remote sensing imagery.
• To investigate the effectiveness of a combined Random Forest Regression and Multiple Linear Regression model for correcting GEDI footprints' overestimation of low-stature vegetation.

Study Configuration

• Spatial Scale: Regional scale, focusing on Qilian Mountain National Park in northwest China.
• Temporal Scale: Data collected from March to October 2022.

Methodology and Data

• Models used:
  • Random Forest Regression
  • Multiple Linear Regression
  • Depth-wise separable convolutional neural network (DS-CNN)
• Data sources:
  • GEDI Level 2A product
  • Sentinel-2 L2A optical imagery
  • Sentinel-1 GRD SAR data
  • FabDEM terrain attributes
  • Annual China Land Cover Dataset (CLCD)

Main Results

• The calibrated height labels improved the characterization of canopy height across forest, shrub, and grassland areas.
• The optimized model achieved a mean absolute error (MAE) of 2.26 m against field measurements, showing strong consistency with airborne LiDAR validation.

Contributions

• This study provides an innovative method for accurately estimating vegetation canopy height in complex mountainous regions using GEDI data calibration and multi-source remote sensing imagery.
• The results demonstrate the effectiveness of a combined Random Forest Regression and Multiple Linear Regression model for correcting GEDI footprints' overestimation of low-stature vegetation.

Funding

• This research was supported by the National Natural Science Foundation of China (Grant No. 42141003) and the Fundamental Research Funds for the Central Universities (Grant No. 2022JCYJQ001).

Citation

@article{Liang2026Vegetation,
  author = {Liang, Huajun and Bie, Qiang and Zhang, Hongwei and Yao, Wenyu},
  title = {Vegetation canopy height retrieval in complex mountainous regions based on data calibration and CNN model},
  journal = {International Journal of Applied Earth Observation and Geoinformation},
  year = {2026},
  doi = {10.1016/j.jag.2026.105598},
  url = {https://doi.org/10.1016/j.jag.2026.105598}
}