Park et al. (2025) Detecting Abandoned Cropland in Monsoon-Influenced Regions Using HLS Imagery and Interpretable Machine Learning

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

Identification

• Journal: Agronomy
• Year: 2025
• Date: 2025-11-24
• Authors: Sinyoung Park, Sanae Kang, Byungmook Hwang, Dongwook W. Ko
• DOI: 10.3390/agronomy15122702

Research Groups

Not explicitly mentioned in the provided text.

Short Summary

This study developed a robust framework combining Harmonized Landsat and Sentinel-2 (HLS) imagery with the XGBoost algorithm to accurately monitor abandoned cropland, achieving an accuracy of 0.84.

Objective

• To establish a robust framework for large-scale quantitative monitoring of abandoned cropland by capturing seasonal spectral variations using high-resolution satellite imagery and machine learning, addressing challenges like frequent cloud cover and fragmented croplands.

Study Configuration

• Spatial Scale: Large-scale, with a spatial resolution of 30 meters.
• Temporal Scale: High temporal resolution (2–3 days), capturing seasonal spectral variations.

Methodology and Data

• Models used: eXtreme Gradient Boosting (XGBoost) algorithm, Balanced Bagging Classifier (BBC), SHapley Additive exPlanations (SHAP) analysis.
• Data sources: Harmonized Landsat and Sentinel-2 (HLS) imagery.

Main Results

• The model achieved an accuracy of 0.84, Cohen’s kappa of 0.71, and an F2 score of 0.84.
• SHAP analysis identified key features including Near-Infrared (NIR) in May–June, Shortwave Infrared 2 (SWIR2) in January, Modified Chlorophyll Absorption Ratio Index (MCARI) in September, and Bare Soil Index (BSI) from January–April, reflecting phenological differences among cropland types.

Contributions

• This study establishes a robust and adaptable framework for large-scale cropland monitoring, overcoming challenges of frequent cloud cover and highly fragmented croplands across diverse regional and climatic conditions, which is a significant advancement for satellite-based detection.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Park2025Detecting,
  author = {Park, Sinyoung and Kang, Sanae and Hwang, Byungmook and Ko, Dongwook W.},
  title = {Detecting Abandoned Cropland in Monsoon-Influenced Regions Using HLS Imagery and Interpretable Machine Learning},
  journal = {Agronomy},
  year = {2025},
  doi = {10.3390/agronomy15122702},
  url = {https://doi.org/10.3390/agronomy15122702}
}