Belaid et al. (2025) Hemispherical imaging of canopy light interception: A ceptometer alternative for precision irrigation in orchards and vineyards

Identification

• Journal: Agricultural and Forest Meteorology
• Year: 2025
• Date: 2025-11-24
• Authors: Mohamed Ibrahim Belaid, Alexandre Escolà Agustí, Jaume Casadesús Brugués
• DOI: 10.1016/j.agrformet.2025.110958

Research Groups

• Efficient Use of Water in Agriculture Program, Institute of AgriFood Research and Technology (IRTA), Fruitcentre, Lleida, Catalonia, Spain
• Research Group in AgroICT & Precision Agriculture, Universitat de Lleida/Agrotecnio-CERCA Center, Lleida, Catalonia, Spain

Short Summary

This study introduces a cost-effective hemispherical imaging method for rapidly and reliably assessing the diurnal pattern and daily fraction of Intercepted Photosynthetically Active Radiation (fIPAR) in orchards and vineyards. The proposed method, utilizing either still or action cameras, demonstrated high agreement with traditional ceptometer measurements (R² between 0.88–0.92) and offers a practical, scalable, and labor-efficient alternative for precision irrigation.

Objective

• To analyze the performance and practical utility of an approach based on hemispherical images for assessing the diurnal curve of fIPAR in tree crops and vines.

Study Configuration

• Spatial Scale: Orchards (apple, almond) and vineyards (Vitis vinifera cv. Chardonnay, Vitis vinifera cv. Pinot) located in Lleida, Catalonia, Spain. The study included various tree spacings (e.g., 1.2 m x 3.6 m for apples, 1.7 m x 2.5 m for vineyards, 3.5 m x 5.5 m for almonds) and tree dimensions.
• Temporal Scale: Diurnal patterns of fIPAR (calculated at 10-minute intervals), daily fIPAR values. Data collection occurred over 2 to 3 consecutive days per site, with measurements spanning from sunrise to sunset.

Methodology and Data

• Models used: Custom-made software developed in Python using the OpenCV library (version 1.6) for automated image processing (canopy segmentation, sun path projection). Solar curve parameters (declination, elevation, azimuth, solar time angle) were calculated based on methodologies from Allen et al. (1998) and Masters (2004).
• Data sources:
  • Hemispherical images:
    • Still camera: Nikon D70 reflex camera equipped with an AT-X 107 DX (Tokina) fisheye zoom lens, mounted on a self-levelling support approximately 20 cm above the ground.
    • Action camera: GoPro HERO10 Black, mounted on a selfie stick and held approximately 20 cm from the ground.
  • Ground truth/comparison data: Instantaneous fIPAR measurements obtained using a portable ceptometer (AccuPAR model LP-80; Decagon Devices Inc.).
  • Calibration data: Lab-based camera calibration using a board with precisely positioned magnets.
  • Geographical data: Latitude, longitude, day of year, and row orientation for solar path calculations.

Main Results

• The hemispherical imaging method showed high agreement with ceptometer measurements for diurnal fIPAR patterns, with R² values ranging from 0.88 to 0.92 and mean square differences of 0.07 to 0.08.
• The method successfully distinguished distinct diurnal fIPAR patterns across different canopy types (e.g., Chardonnay-large vs. Chardonnay-small).
• High repeatability of the method was observed across varying lighting conditions (dawn vs. daylight), with R² values consistently greater than 0.98.
• fIPAR values calculated from daylight images were slightly lower (3%) than those from dawn images, primarily attributed to the effect of sun flares. The maximum potential error from sun flares was quantified as less than 3% of analyzed pixels.
• Action cameras significantly reduced field sampling time (1 to 2 minutes per canopy) compared to still cameras (10 to 12 minutes) and ceptometers (8 to 10 minutes).
• Action cameras performed better for tall, dense canopies (almond, apple) but tended to overestimate fIPAR in shorter, more open canopies (vineyards) due to a wider field of view.
• The canopy segmentation algorithm was robust under various background conditions (clear sky, clouds, grey anti-hail nets), requiring only minor adjustments for colored (yellow) anti-hail nets.

Contributions

• Proposes and validates a novel, cost-effective, and labor-efficient hemispherical imaging methodology for assessing the diurnal pattern and daily integrated fIPAR in orchards and vineyards, overcoming the limitations of single midday ceptometer measurements.
• Demonstrates the operational viability of using both still cameras (for rigorous characterization) and action cameras (for rapid, scalable, and cost-effective data acquisition) as practical alternatives to traditional fIPAR measurement tools.
• Shows the method's robustness and adaptability across diverse lighting conditions (dawn, daylight, overcast) and various canopy architectures, including those under anti-hail nets.
• Provides a valuable tool for precision irrigation and orchard management, enabling more comprehensive canopy structure assessments and facilitating comparisons between orchards with different training systems.

Funding

• Project RTA2010-00007-00-00 (Spanish Ministry of Agriculture)
• Project ET4DROUGHT, PID2021-127345OR-C31 (Spanish Agencia Espacial de Investigación)
• Joint fellowship from University of Lleida and IRTA (for M.I.B)

Citation

@article{Belaid2025Hemispherical,
  author = {Belaid, Mohamed Ibrahim and Agustí, Alexandre Escolà and Brugués, Jaume Casadesús},
  title = {Hemispherical imaging of canopy light interception: A ceptometer alternative for precision irrigation in orchards and vineyards},
  journal = {Agricultural and Forest Meteorology},
  year = {2025},
  doi = {10.1016/j.agrformet.2025.110958},
  url = {https://doi.org/10.1016/j.agrformet.2025.110958}
}