Dong et al. (2025) Integrating prior information for improving 3D model-driven GAI estimation with application to wheat crops

Identification

• Journal: Remote Sensing of Environment
• Year: 2025
• Date: 2025-11-29
• Authors: Mingxia Dong, Shouyang Liu, Marie Weiss, Aojie Yin, Chen Zhu, Benoît de Solan, Wei Guo, Fernandes Richard, Wenjuan Li, Xia Yao, James Burridge, Zhen Chen, Yanfeng Ding, Frédéric Baret
• DOI: 10.1016/j.rse.2025.115161

Research Groups

• Engineering Research Center of Plant Phenotyping, Ministry of Education, State Key Laboratory of Crop Genetics & Germplasm Enhancement and Utilization, Jiangsu Collaborative Innovation Center for Modern Crop Production, Academy for Advanced Interdisciplinary Studies, Nanjing Agricultural University, Nanjing, China
• EMMAH UMR 1114, INRAE, Domaine Saint-Paul, Site Agroparc, Avignon, France
• ARVALIS Institut du végétal, Paris, France
• Graduate School of Agricultural and Life Sciences, The University of Tokyo, Tokyo, Japan
• Canada Centre for Remote Sensing, Natural Resources Canada, Ottawa, Canada
• State Key Laboratory of Efficient Utilization of Arid and Semi-arid Arable Land in Northern China, The Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing, China
• National Engineering and Technology Center for Information Agriculture (NETCIA), MARA Key Laboratory of Crop System Analysis and Decision Making, MOE, Engineering Research Center of Smart Agriculture, Jiangsu Key Laboratory for Information Agriculture, Institute of Smart Agriculture, Nanjing Agricultural University, Nanjing, China
• Institute of Farmland Irrigation, Chinese Academy of Agricultural Sciences, Xinxiang, China

Short Summary

This study investigated the integration of prior information (soil background, leaf optical properties, and canopy structure) into radiative transfer models to enhance Green Area Index (GAI) estimation for wheat crops. It demonstrated that stage-specific GAI retrieval with detailed prior information significantly improves accuracy compared to standard model inversion approaches.

Objective

• To comprehensively explore how to integrate prior information (soil background, leaf optical properties, and canopy structure) into radiative transfer models to improve Green Area Index (GAI) retrieval for wheat crops.

Study Configuration

• Spatial Scale: Sub-10 meter resolution (SuperDove satellite data).
• Temporal Scale: Stage-specific monitoring across crop growth stages.

Methodology and Data

• Models used: MARMIT-2 (soil reflectance), PROSPECT (leaf optical properties), ADEL-Wheat (canopy structure), LESS (radiative transfer), Support Vector Regression (SVR).
• Data sources: SuperDove satellite multispectral reflectance observations, 310 samples of GAI ground measurements.

Main Results

• Stage-specific GAI retrieval integrating detailed prior information on soil and leaf properties achieved significantly higher accuracy (R² = 0.93, Root Mean Square Error (RMSE) = 0.47) compared to standard model inversion approaches (R² = 0.82, RMSE = 0.73).
• The improved realism of the training dataset was attributed to three key strategies: (1) employing models that integrate physical and biological knowledge; (2) narrowing the training space; and (3) minimizing distribution shifts.

Contributions

• Demonstrates a notable improvement in GAI estimation accuracy for wheat crops by effectively integrating detailed prior information into 3D model-driven approaches.
• Provides a comprehensive framework for leveraging physical and biological knowledge through a suite of radiative transfer models to generate more realistic training datasets.
• Identifies and discusses key strategies (integrating physical/biological knowledge, narrowing training space, minimizing distribution shifts) for enhancing the realism and performance of model-driven GAI retrieval.
• Suggests the methodology and findings are extensible to other crop types, vegetation variables, and satellite remote sensing systems.

Funding

Not explicitly mentioned in the provided paper text.

Citation

@article{Dong2025Integrating,
  author = {Dong, Mingxia and Liu, Shouyang and Weiss, Marie and Yin, Aojie and Zhu, Chen and Solan, Benoît de and Guo, Wei and Richard, Fernandes and Li, Wenjuan and Yao, Xia and Burridge, James and Chen, Zhen and Ding, Yanfeng and Baret, Frédéric},
  title = {Integrating prior information for improving 3D model-driven GAI estimation with application to wheat crops},
  journal = {Remote Sensing of Environment},
  year = {2025},
  doi = {10.1016/j.rse.2025.115161},
  url = {https://doi.org/10.1016/j.rse.2025.115161}
}