Kong et al. (2026) Integrating deep-learning detection and hydrological monitoring for driftwood flux estimation in a steep tropical watershed during a typhoon-driven flood

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-08-06
• Authors: Qian-Yun Kong, Ci‐Jian Yang, Cheng-Hua Tsai, Meng-Yu Lee
• DOI: 10.1016/j.jhydrol.2026.136183

Research Groups

Department of Geography, National Taiwan University, Taipei, Taiwan

Short Summary

This study integrates deep-learning detection with hydrological monitoring to estimate driftwood flux in a steep tropical watershed during a typhoon-driven flood.

Objective

• Investigate the feasibility of using deep learning-based detection for quantifying event-scale driftwood flux and storage in steep mountain watersheds.

Study Configuration

• Spatial Scale: A steep tropical watershed in Taiwan.
• Temporal Scale: During a typhoon-driven flood event (Typhoon Kong-rey, 2024).

Methodology and Data

• Models used: YOLOv8 for object detection and instance segmentation.
• Data sources: Fixed-interval surveillance cameras, drone imagery, and RFID-tagged logs.

Main Results

• Estimated instream driftwood mass of 1.12 ± 0.22 × 10^4 t and a corresponding carbon stock of 5.26 ± 1.05 × 10^3 tC during Typhoon Kong-rey.
• Driftwood preferentially accumulates along outer river banks and forms banded deposits on nearshore beaches.

Contributions

• Demonstrated the feasibility of integrating deep learning-based detection with hydrological monitoring for carbon budget assessments in steep tropical watersheds.

Funding

• This research was funded by [project/program name and reference code, e.g., "National Science Council" or "NSC 104-2625-M-002-001"].

Citation

@article{Kong2026Integrating,
  author = {Kong, Qian-Yun and Yang, Ci‐Jian and Tsai, Cheng-Hua and Lee, Meng-Yu},
  title = {Integrating deep-learning detection and hydrological monitoring for driftwood flux estimation in a steep tropical watershed during a typhoon-driven flood},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136183},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136183}
}