Lin et al. (2025) Pluvial flood warnings in a tidally influenced basin enhanced by rainfall–tide compound indicators: Taipei case study

Identification

• Journal: Stochastic Environmental Research and Risk Assessment
• Year: 2025
• Date: 2025-11-20
• Authors: M. Lin, Wen-Yih Sun, Huei-Syuan Fu, Chih-Hsin Chang, Wei‐Bo Chen, Chieh-Ju Wang, Hengxi Wang, Yi‐Chiang Yu, Y.S. Li
• DOI: 10.1007/s00477-025-03120-y

Research Groups

• Slopeland and Hydrology Division, National Science and Technology Center for Disaster Reduction, New Taipei City, Taiwan
• Meteorology Division, National Science and Technology Center for Disaster Reduction, New Taipei City, Taiwan
• Climate Change Division, National Science and Technology Center for Disaster Reduction, New Taipei City, Taiwan
• Department of Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, IN, USA
• Department of Atmospheric Sciences, National Central University, Zhongli, Taiwan

Short Summary

This study developed and evaluated rainfall-tide compound indicators using a copula-based statistical framework to enhance pluvial flood warnings in the tidally influenced Taipei metropolitan area. The compound indicators significantly improved flood detection rates (up to 84.3%) compared to rainfall-only thresholds, highlighting the critical role of tidal interaction in urban flood dynamics.

Objective

• Describe the hydrological and urban context of the Tamsui River Basin.
• Develop an improved compound flood index incorporating 3-hour rainfall lead time using copula theory.
• Evaluate the predictive performance of the compound flood index across flood-prone urban districts.
• Discuss implications for early warning systems and urban governance in tidal metropolitan areas.

Study Configuration

• Spatial Scale: Taipei metropolitan area (Taipei City and New Taipei City), specifically 11 districts in Taipei City and 10 districts in New Taipei City, within the Tamsui River Basin, northern Taiwan.
• Temporal Scale: Urban flood events recorded between 2001 and 2024. Rainfall accumulation periods of 1 hour and 3 hours.

Methodology and Data

• Models used:
  • Copula-based statistical framework (Gaussian copula).
  • Empirical urban flood risk model (based on river discharge principles).
  • Compound flooding indicator equations: $T(t)=C \cdot D(t)+\delta$.
• Data sources:
  • Flood incident records (approximately 4000 reports from Taiwan's Central Emergency Operation Center).
  • Rainfall measurements (district-level gauge stations).
  • Tidal levels (Tamsui tide gauge station, incorporating astronomical tides and typhoon storm surges).
  • Flood depth sensor data.

Main Results

• The Gaussian copula was identified as the optimal model for capturing the dependence structure between rainfall and tide (Spearman ρ = − 0.433; Pearson ρ = − 0.392).
• Compound indicators substantially outperformed rainfall-only thresholds in flood detection.
• In Taipei City, detection rates increased from 59% to 82.2% for 1-hour rainfall and from 72% to 82% for 3-hour rainfall.
• In New Taipei City, the 3-hour compound indicator improved detection from 64.6% to 84.3%.
• Compound flooding is most likely when moderate-to-high values of rainfall and tide co-occur (peak probability densities around rainfall CDFs of 0.6–0.8 and tidal CDFs of 0.5–0.7).
• The 3-hour compound indicator showed superior sensitivity in capturing flood onset and persistence in low-lying, slow-draining areas and aligned with observed flooding during Typhoon Nesat (2022) where 1-hour indicators failed.

Contributions

• Developed an improved copula-based compound flood indicator that integrates rainfall accumulation (1-hour and 3-hour) and real-time tidal levels, specifically tailored for tidally influenced urban environments.
• Demonstrated the critical role of tidal interaction in urban flood dynamics, showing that compound indicators significantly enhance flood detection accuracy compared to traditional rainfall-only thresholds.
• Provided a practical and operationally viable tool for urban flood early warning systems in tidal river basins, capable of incorporating forecast uncertainty.
• Extended previous work by incorporating a 3-hour rainfall accumulation period, improving predictive capability for scenarios involving runoff accumulation and drainage delays.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Citation

@article{Lin2025Pluvial,
  author = {Lin, M. and Sun, Wen-Yih and Fu, Huei-Syuan and Chang, Chih-Hsin and Chen, Wei‐Bo and Wang, Chieh-Ju and Wang, Hengxi and Yu, Yi‐Chiang and Li, Y.S.},
  title = {Pluvial flood warnings in a tidally influenced basin enhanced by rainfall–tide compound indicators: Taipei case study},
  journal = {Stochastic Environmental Research and Risk Assessment},
  year = {2025},
  doi = {10.1007/s00477-025-03120-y},
  url = {https://doi.org/10.1007/s00477-025-03120-y}
}