Zhao et al. (2025) Evaluating the impact of canopy spatial heterogeneity on Solar-Induced chlorophyll fluorescence and Model-Based quantification

Identification

• Journal: International Journal of Applied Earth Observation and Geoinformation
• Year: 2025
• Date: 2025-11-20
• Authors: Liang Zhao, Rui Sun
• DOI: 10.1016/j.jag.2025.104971

Research Groups

• State Key Laboratory of Remote Sensing and Digital Earth, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
• Beijing Engineering Research Center for Global Land Remote Sensing Products, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
• Faculty of Arts and Sciences, Beijing Normal University, Zhuhai 519087, China

Short Summary

This study quantifies the impact of canopy spatial heterogeneity on Solar-Induced chlorophyll fluorescence (SIF) using a 3D radiative transfer model. It develops a correction function based on vegetation index-derived coefficients of variation (CVs) to convert homogeneous-canopy SIF to heterogeneous-canopy conditions, thereby improving SIF estimation accuracy.

Objective

• To assess the differences in SIF between heterogeneous and homogeneous scenes and analyze changes in angular dependency after estimating total leaf-emitted SIF (SIFtotal) from top-of-canopy SIF (SIFtoc).
• To compare the sensitivity differences of SIFtotal and SIFtoc with respect to the fraction of absorbed photosynthetically active radiation (fAPAR) and near-infrared reflectance of vegetation (NIRv).
• To develop a method for correcting homogeneous scene SIF to true SIF based on canopy heterogeneity indices.

Study Configuration

• Spatial Scale: Simulated 3D forest scenes of 30 meters × 30 meters; vegetation index (VI) data aggregated to 1 meter, 5 meters, and 10 meters resolutions. Real-world reference data from Saihanba Forest Farm, China, using Sentinel-2 (10 meters resolution for relevant bands) and GLASS LAI (250 meters resolution).
• Temporal Scale: Sentinel-2 data from June 21, 2020; GLASS LAI product for day 169 of 2020. The study is primarily a simulation-based analysis of static canopy structures.

Methodology and Data

• Models used:
  • 3D discrete anisotropic radiative transfer (DART) model (DART-FT method) for SIF simulation, radiation budget, and bidirectional reflectance factor (BRF).
  • Fluspect-Cx model, integrated into DART, for simulating leaf optical properties and fluorescence excitation-emission matrices.
• Data sources:
  • 18 simulated 3D forest scenes, including various tree distribution patterns (Larch, Birch) and ground configurations (grassland), constructed using OnyxTREE software.
  • Soil spectra ("sandyloambrown") and birch wood spectral data ("betulapendulabran") from the DART database.
  • Larch wood spectral data from Hovi et al. (2017).
  • Sentinel-2 satellite data (red and near-infrared bands) for real-world validation of NIRv.
  • ESA WorldCover 2020 dataset for forest distribution data.
  • GLASS LAI product (250 meters resolution) for LAI reference in real-world validation.

Main Results

• Both SIFtotal and hemispherically integrated top-of-canopy SIF (SIFhemi) are significantly higher in homogeneous canopies than in heterogeneous canopies. For LAI = 3, SIFtotal in homogeneous scenes exceeded heterogeneous scenes by 22.13 %–56.03 % (640–850 nm), 22.47 %–56.96 % (685 nm), and 22.07 %–55.88 % (740 nm).
• Estimating SIFtotal from SIFtoc effectively reduces angular dependency, with angular variability of estimated SIFtotal (AVSIFtotal) ranging from −13.60 % to 21.63 %, a significant improvement compared to AVSIFtoc (ranging from –33.87 % to 78.39 %).
• SIFtotal exhibits a very strong linear correlation with fAPAR (R² = 1 in all three wavelength bands: 640–850 nm, 685 nm, and 740 nm), while SIFtoc (nadir) shows stronger correlations with NIRv (R² = 0.99 at 740 nm) and EVI (R² = 0.98 at 740 nm).
• Vegetation index-derived coefficients of variation (CVs) are negatively correlated with SIF, indicating that greater canopy heterogeneity leads to reduced SIF. CVNIRv and CVEVI at 10 meters resolution demonstrated robust performance in fitting the ratio of heterogeneous-to-homogeneous canopy SIF, validating the proposed correction function.

Contributions

• Quantified the significant impact of canopy structural heterogeneity on SIFtotal and SIFhemi, highlighting the biases introduced by homogeneous-canopy assumptions.
• Demonstrated that converting SIFtoc to SIFtotal effectively reduces angular dependence, improving the reliability of SIF estimates.
• Clarified the distinct relationships between SIF types and canopy structural parameters: SIFtotal is strongly linked to fAPAR, while SIFtoc is better correlated with NIRv and EVI.
• Introduced a practical correction framework using VI-derived CVs (specifically CVNIRv and CVEVI at 10 meters resolution) to translate homogeneous-canopy SIF simulations to heterogeneous-canopy conditions, thereby bridging the gap between 1D and 3D radiative transfer models and enhancing SIF estimation accuracy for large-scale applications.

Funding

• National Natural Science Foundation of China (42271330)

Citation

@article{Zhao2025Evaluating,
  author = {Zhao, Liang and Sun, Rui},
  title = {Evaluating the impact of canopy spatial heterogeneity on Solar-Induced chlorophyll fluorescence and Model-Based quantification},
  journal = {International Journal of Applied Earth Observation and Geoinformation},
  year = {2025},
  doi = {10.1016/j.jag.2025.104971},
  url = {https://doi.org/10.1016/j.jag.2025.104971}
}