Zhao et al. (2025) Optimizing Cloud Mask Accuracy over Snow-Covered Terrain with a Multistage Decision Tree Framework

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

Identification

• Journal: Remote Sensing
• Year: 2025
• Date: 2025-12-10
• Authors: Qin Zhao, Xiaohua Hao, Donghang Shao, Wenzheng Ji, Guanghui Huang, Zisheng Zhao, Juan Zhang
• DOI: 10.3390/rs17243992

Research Groups

Not available from the provided text.

Short Summary

This study developed an enhanced multi-threshold cloud detection algorithm using AVHRR data to improve cloud-snow discrimination in optical remote sensing, achieving an overall accuracy of 82.08% and significantly outperforming existing methods, especially in snow-covered mountainous regions.

Objective

• To develop an enhanced multi-threshold cloud detection algorithm to overcome cloud obstruction and spectral confusion between clouds and snow in high-resolution optical remote sensing imagery, thereby improving data quality and application reliability.

Study Configuration

• Spatial Scale: Earth's surface, with a particular focus on high-altitude mountainous regions with snow cover, utilizing high-resolution imagery.
• Temporal Scale: Continuous monitoring (implied by the need to address persistent cloud overestimation).

Methodology and Data

• Models used: Enhanced multi-threshold cloud detection algorithm incorporating dynamic threshold optimization within a multi-level decision tree framework.
• Data sources: AVHRR surface reflectance data (for algorithm development and application); Landsat 5 SR (as reference data for validation).

Main Results

• The developed algorithm achieved an overall accuracy (OA) of 82.08% for cloud-snow discrimination.
• User's accuracy (UA) reached 79.41%, and the F-score was 82.55%.
• The proposed algorithm outperformed two existing algorithms, showing OA improvements of 17.42% and 7.93%, respectively.
• A significant reduction in cloud misidentification was observed, with UA increases of 21.02% and 13.21% compared to existing methods.
• These improvements were most pronounced in high-altitude mountainous regions with snow cover.
• The algorithm maintains computational efficiency while providing reliable cloud masking.

Contributions

• Development of a novel enhanced multi-threshold cloud detection algorithm featuring dynamic threshold optimization and a multi-level decision tree framework.
• Significant improvement in cloud-snow discrimination accuracy, particularly reducing cloud misidentification in challenging snow-covered mountainous regions.
• Provides a computationally efficient and reliable cloud masking solution, enhancing support for snow cover monitoring and broader environmental applications.

Funding

Not available from the provided text.

Citation

@article{Zhao2025Optimizing,
  author = {Zhao, Qin and Hao, Xiaohua and Shao, Donghang and Ji, Wenzheng and Huang, Guanghui and Zhao, Zisheng and Zhang, Juan},
  title = {Optimizing Cloud Mask Accuracy over Snow-Covered Terrain with a Multistage Decision Tree Framework},
  journal = {Remote Sensing},
  year = {2025},
  doi = {10.3390/rs17243992},
  url = {https://doi.org/10.3390/rs17243992}
}