Guo et al. (2025) An improved gap probability estimation method accounting for radiometric effects in airborne LiDAR intensity

Identification

• Journal: International Journal of Applied Earth Observation and Geoinformation
• Year: 2025
• Date: 2025-12-10
• Authors: Lijie Guo, Lei Deng
• DOI: 10.1016/j.jag.2025.105010

Research Groups

• College of Resource Environment and Tourism, Capital Normal University, Beijing, China

Short Summary

This paper introduces PRE_COR, a novel method for estimating vegetation canopy gap probability (P) from airborne LiDAR intensity data by correcting for radiometric effects (distance and incidence angle). The method significantly enhances P estimation accuracy and stability, particularly under large scan angle conditions, outperforming traditional approaches.

Objective

• To develop and validate an improved method (PRE_COR) for estimating vegetation canopy gap probability (P) from airborne LiDAR intensity data by explicitly correcting for radiometric effects, such as sensor-to-target distance and incidence angle.

Study Configuration

• Spatial Scale: Simulated scenes of 30 m × 30 m; realistic data processed in 30 m × 30 m plots within 1 km × 1 km tiles.
• Temporal Scale: NEON data acquired on June 18, 2023 (SCBI site) and August 4, 2022 (HARV site).

Methodology and Data

• Models used:
  • PRE_COR (Proposed method for Gap Probability Estimation Accounting for Radiometric Effects in Airborne LiDAR Intensity)
  • PFitted (Traditional intensity-based method for comparison)
  • LESS (Large-scale remote sensing data and image simulation model)
  • RPV (Rahman–Pinty–Verstraete model for non-Lambertian canopy reflectance)
  • SOILSPECT (Model for non-Lambertian soil reflectance)
  • Inverse distance square law and Cosine law of incidence angle (for radiometric correction)
• Data sources:
  • Simulated airborne laser scanning (ALS) point cloud data (generated using LESS).
  • Simulated four-component images (for reference P in virtual scenes).
  • National Ecological Observatory Network (NEON) ALS point cloud data (Optech ALTM Gemini system, 1064 nm wavelength).
  • NEON high-resolution RGB orthorectified imagery (D8900 digital camera, 0.1 m pixel size, for reference P in realistic scenes).

Main Results

• Simulated Data:
  • PRECOR consistently outperformed the traditional PFitted method across all scenarios.
  • Under varying canopy cover, PRECOR reduced relative root mean square error (rRMSE) by up to 1.90% and mean absolute error (MAE) by up to 0.010.
  • Under varying flight altitudes, PRECOR reduced rRMSE by up to 1.94% and MAE by up to 0.003.
  • Under varying mean scan angles, PRECOR significantly reduced rRMSE by at least 17.11% (up to 21.20%) and MAE by at least 0.067 (up to 0.096), with benefits being most pronounced at larger scan angles.
• NEON ALS Data (Realistic):
  • Overall, PRECOR improved the coefficient of determination (R²) by 0.030 to 0.058, reduced rRMSE by 2.36% to 4.02%, and decreased MAE by 0.006 to 0.009 compared to PFitted.
  • For scan angles exceeding 20°, PRECOR showed rRMSE improvements potentially exceeding 5.39% and MAE improvements potentially exceeding 0.019.
• Non-Lambertian Effects: Deviations in P between Lambertian and non-Lambertian assumptions were small (MAE between 0.008 and 0.033), supporting the reasonableness of the Lambertian assumption for this study.

Contributions

• Proposes a novel method (PRE_COR) for canopy gap probability (P) estimation that explicitly corrects for radiometric effects (distance and incidence angle) in airborne LiDAR intensity data.
• Demonstrates that radiometric correction significantly enhances the accuracy and stability of P estimation, particularly under challenging conditions such as large scan angles.
• Provides a robust and reliable approach for P estimation, which is critical for accurate Leaf Area Index (LAI) retrieval and subsequent ecological modeling (e.g., photosynthesis, evapotranspiration, carbon cycling).
• Validates the method using a comprehensive set of simulated data (varying canopy cover, flight altitudes, mean scan angles, and crown shapes) and real-world NEON ALS data from two distinct forest sites.

Funding

• R&D program of Beijing Municipal Education Commission [grant number No. KZ202210028045].

Citation

@article{Guo2025improved,
  author = {Guo, Lijie and Deng, Lei},
  title = {An improved gap probability estimation method accounting for radiometric effects in airborne LiDAR intensity},
  journal = {International Journal of Applied Earth Observation and Geoinformation},
  year = {2025},
  doi = {10.1016/j.jag.2025.105010},
  url = {https://doi.org/10.1016/j.jag.2025.105010}
}