Hu et al. (2025) Evaluation of UAV and satellite platforms for gross primary production monitoring under different model frameworks in agroecosystems

Identification

• Journal: Computers and Electronics in Agriculture
• Year: 2025
• Date: 2025-11-13
• Authors: Xiaolong Hu, Jiang Bian, Liangsheng Shi, Lin Lin, Shenji Li, Xianzhi Deng, Jinmin Li, Chenye Su, Shuai Du, Tinghan Wang, Yujie Wang
• DOI: 10.1016/j.compag.2025.111182

Research Groups

• State Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan, Hubei 430072, China
• College of Water Resources and Architectural Engineering, Northwest A & F University, Yangling 712100 Shaanxi, China
• Urban Operation Management Center of Hengsha Township, Shanghai 201914, China

Short Summary

This study evaluated the performance of Unmanned Aerial Vehicle (UAV) and satellite platforms (Sentinel-2, MODIS) for monitoring gross primary production (GPP) in agroecosystems under various model frameworks, finding that UAVs provide more accurate GPP estimates, particularly with simpler models.

Objective

• To assess the performance of Unmanned Aerial Vehicle (UAV) and satellite platforms (Sentinel-2, MODIS) for monitoring gross primary production (GPP) under the frameworks of Boreal Ecosystem Productivity Simulator (BEPS), light use efficiency (LUE), and linear regression (LR) based statistical models using three agroecosystem eddy covariance sites.

Study Configuration

• Spatial Scale: Three agroecosystem eddy covariance sites (planted with maize, wheat, and rice, respectively).
• Temporal Scale: Crop growing seasons at three agroecosystem sites.

Methodology and Data

• Models used: Boreal Ecosystem Productivity Simulator (BEPS), Light Use Efficiency (LUE) model, Linear Regression (LR) based statistical model.
• Data sources: Unmanned Aerial Vehicle (UAV), Sentinel-2 satellite, Moderate Resolution Imaging Spectroradiometer (MODIS) satellite, Eddy covariance (EC) system (for ground truth/validation).

Main Results

• UAV platforms yield more accurate GPP estimates due to their high-quality data compared to Sentinel-2 and MODIS.
• The root mean square error (RMSE) of UAV-based GPP estimates ranged from 6 to 13.22 µmol m⁻² s⁻¹.
• The RMSE of Sentinel-2 and MODIS-based GPP estimates increased by 8–47 % and 10–42 %, respectively, compared to UAV.
• The advantage of UAV over satellites is particularly pronounced when using LR and LUE models, which have simpler structures.
• LR and LUE models generally outperform the BEPS model, likely due to a strong linear relationship between GPP and solar radiation at the field scale.

Contributions

• Provides a comprehensive comparative assessment of UAV and satellite remote sensing platforms (Sentinel-2, MODIS) for GPP monitoring across different model frameworks (BEPS, LUE, LR) in agroecosystems.
• Quantifies the superior accuracy of UAV-based GPP estimates and highlights its advantages over satellite platforms, especially with simpler GPP models.
• Offers valuable guidance for selecting appropriate remote sensing platforms and model frameworks to achieve accurate GPP monitoring in agroecosystems.

Funding

• Not specified in the provided text.

Citation

@article{Hu2025Evaluation,
  author = {Hu, Xiaolong and Bian, Jiang and Shi, Liangsheng and Lin, Lin and Li, Shenji and Deng, Xianzhi and Li, Jinmin and Su, Chenye and Du, Shuai and Wang, Tinghan and Wang, Yujie},
  title = {Evaluation of UAV and satellite platforms for gross primary production monitoring under different model frameworks in agroecosystems},
  journal = {Computers and Electronics in Agriculture},
  year = {2025},
  doi = {10.1016/j.compag.2025.111182},
  url = {https://doi.org/10.1016/j.compag.2025.111182}
}