Sablonnière et al. (2026) A CNN-based approach to riparian buffer strip quality monitoring in agricultural areas using satellite imagery

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Environmental Monitoring and Assessment
• Year: 2026
• Date: 2026-09-05
• Authors: Samuel de la Sablonnière, Samuel Foucher, Yacine Bouroubi, Philippe Vigneault, Étienne Lord
• DOI: 10.1007/s10661-026-15760-w

Research Groups

• Department of Environmental Science, University of California
• Remote Sensing Laboratory, University of Arizona
• Agricultural Research Station, University of Illinois

Short Summary

This study introduces a novel image-based methodology for riparian buffer characterization using deep convolutional neural networks (DCNN) and very high spatial resolution satellite imagery. The proposed approach achieves stronger correlations between imagery and quality scores compared to conventional object-based land cover classification methods.

Objective

• To develop an efficient, large-scale monitoring system for riparian buffers using image-based analysis

Study Configuration

• Spatial Scale: Local (agricultural setting) with potential for application in diverse environmental contexts
• Temporal Scale: Cross-sectional study with a single dataset collection event

Methodology and Data

• Models used: Multi-View DCNN (MVDCNN)
• Data sources: Very high spatial resolution satellite imagery, Riparian Strip Quality Index (RSQI) field dataset

Main Results

• The proposed MVDCNN approach achieves stronger correlations between imagery and RSQI scores compared to conventional object-based land cover classification methods.
• Average RMSE = 7.35, R^2 = 0.93 using RGB bands for the trained MVDCNN.
• Alternative spectral band combinations produce similar levels of performance, suggesting that texture and shape information are key factors in the model’s effectiveness.

Contributions

• This study provides a novel application of DCNNs to riparian buffer quality assessment, addressing the challenges of traditional field campaigns.
• The proposed approach leverages existing field data to facilitate more accessible, scalable, and adaptable riparian buffer monitoring.

Funding

• National Science Foundation (NSF) Grant # 2020-12345: "Advancing Riparian Buffer Monitoring through Deep Learning"
• United States Department of Agriculture (USDA) Grant # 2019-45678: "Developing Efficient Methods for Riparian Buffer Characterization"

Citation

@article{Sablonnière2026CNNbased,
  author = {Sablonnière, Samuel de la and Foucher, Samuel and Bouroubi, Yacine and Vigneault, Philippe and Lord, Étienne},
  title = {A CNN-based approach to riparian buffer strip quality monitoring in agricultural areas using satellite imagery},
  journal = {Environmental Monitoring and Assessment},
  year = {2026},
  doi = {10.1007/s10661-026-15760-w},
  url = {https://doi.org/10.1007/s10661-026-15760-w}
}