Alizadeh et al. (2025) Improving crop biophysical parameter estimation using high-resolution multispectral UAV imagery and PROSAIL model

Identification

• Journal: Smart Agricultural Technology
• Year: 2025
• Date: 2025-12-30
• Authors: Hadi Alizadeh, Zahra Azizi, Ali Asghar Alesheikh, Hossein Aghamohammadi, Hamid Reza Pourghasemi
• DOI: 10.1016/j.atech.2025.101772

Research Groups

• Department of Remote Sensing and GIS, SR.C., Islamic Azad University, Tehran, Iran
• Department of Geoinformation and Geomatics Engineering, Faculty of Civil, Water and Environmental Engineering, Shahid Beheshti University, Tehran, Iran
• Department of Geospatial Information Systems, Faculty of Geodesy and Geomatics Engineering, K. N. Toosi University of Technology, Tehran, Iran
• Department of Civil Engineering, SR.C., Islamic Azad University, Tehran, Iran
• Department of Soil Science, College of Agriculture, Shiraz University, Shiraz, Iran

Short Summary

This study evaluated a practical workflow coupling high-resolution unmanned aerial system (UAS) multispectral imagery with PROSAIL inversion to map rice canopy traits across three phenological stages, demonstrating superior accuracy compared to Sentinel-2 for field-scale biophysical parameter estimation.

Objective

• To evaluate a practical workflow using high-resolution unmanned aerial system (UAS) multispectral imagery coupled with PROSAIL inversion to map rice canopy traits across three phenological stages, and to benchmark its performance against Sentinel-2 data.

Study Configuration

• Spatial Scale: Approximately 50 hectares of rice fields in Mazandaran Province, Iran. UAV imagery had ground sampling distances (GSD) ranging from 3.7 cm to 13.06 cm, while Sentinel-2 imagery had GSDs of 10 m, 20 m, and 60 m.
• Temporal Scale: Data collected across three key phenological stages of rice growth in 2020: transplanting (June 10), tillering/peak greenness (July 11), and full maturity/pre-harvest (August 02).

Methodology and Data

• Models used: PROSAIL (PROSPECT + SAIL) radiative transfer model, inverted using a Look-Up Table (LUT) method.
• Data sources:
  • High-resolution multispectral UAV imagery (eBee with Sequoia sensor: Green, Red, Red Edge, Near-Infrared bands).
  • RGB UAV imagery (Phantom 4 Pro V2.0).
  • Sentinel-2 satellite images (13 spectral bands).
  • Independent ground measurements for validation: Leaf Area Index (LAI-2200C), leaf chlorophyll content (SPAD meter), and leaf samples for biochemical parameters (Cab, Car, Cw, Cm, N) from 200 stratified random samples.

Main Results

• The "UAS multispectral + PROSAIL" pipeline accurately retrieved rice biophysical parameters (Cab, Car, Cm, Cw, LAI, N) with strong agreement to field data (R² > 0.98 for most variables at the first date).
• Consistent phenological trajectories were observed: Cab and LAI peaked at maximum greenness, while SIPI2 rose during senescence alongside declining Cab and Cw.
• PROSAIL RMSE maps showed stage-dependent errors, with the lowest and most homogeneous errors detected at peak canopy. For Cab, R²/RMSE values were 0.996/1.555, 0.978/2.104, and 0.972/0.2 across the three dates, respectively.
• Sentinel-2 data yielded lower accuracy at field scale (e.g., LAI R²/RMSE ≈0.81/0.7; Cab ≈0.78/6.5) compared to UAV data, primarily due to coarser resolution and cloud cover constraints.
• The UAV-based approach demonstrated superior performance for precise crop monitoring and within-field mapping of rice biophysics, outperforming Sentinel-2-based estimates.

Contributions

• Demonstrated a highly accurate and practical workflow for field-scale retrieval of rice biophysical and biochemical parameters using high-resolution UAV multispectral imagery and PROSAIL inversion.
• Provided a robust assessment of the temporal dynamics of rice canopy traits across key phenological stages, capturing coherent phenological trajectories.
• Benchmarked the performance of UAV-based retrievals against Sentinel-2, highlighting the superior accuracy and detail of UAV data for within-field applications, especially under atmospheric and management constraints.
• Showcased the enhanced robustness achieved when red-edge and pigment-ratio indices are fused with physical inversion for phenology-dependent retrievals.
• Proposed an operational pathway for precision agriculture applications, such as variable-rate fertilization and irrigation scheduling, through fine-scale, within-field monitoring.

Funding

This study did not receive any grants from public or commercial funding agencies, funds, or any other support.

Citation

@article{Alizadeh2025Improving,
  author = {Alizadeh, Hadi and Azizi, Zahra and Alesheikh, Ali Asghar and Aghamohammadi, Hossein and Pourghasemi, Hamid Reza},
  title = {Improving crop biophysical parameter estimation using high-resolution multispectral UAV imagery and PROSAIL model},
  journal = {Smart Agricultural Technology},
  year = {2025},
  doi = {10.1016/j.atech.2025.101772},
  url = {https://doi.org/10.1016/j.atech.2025.101772}
}