Shao et al. (2025) Bridging Uncertainty in SWMM Model Calibration: A Bayesian Analysis of Optimal Rainfall Selection

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

Identification

• Journal: Water
• Year: 2025
• Date: 2025-12-03
• Authors: Zhiyu Shao, Jinsong Wang, Xiaoyuan Zhang, Jiahui Du, Scott A. Yost
• DOI: 10.3390/w17233435

Research Groups

Not specified in the provided text.

Short Summary

This study establishes a Bayesian SWMM calibration framework to investigate how different rainfall types influence the uncertainty of urban hydrological model parameters, finding that higher intensity, one-year return period rainfall events and double-peak patterns generally yield more accurate and less uncertain parameter estimations.

Objective

• To investigate the influences of rainfall types (intensity and pattern) on the uncertainty of Stormwater Management Model (SWMM) hydrological parameter calibration using Bayesian inference theory.

Study Configuration

• Spatial Scale: Methodological study applicable to urban catchments and watershed management.
• Temporal Scale: Focus on individual rainfall events (e.g., one-year return period) for calibration.

Methodology and Data

• Models used: SWMM (Stormwater Management Model), DREAM(zs) (Differential Evolution Adaptive Metropolis, Version ZS) for Bayesian sampling.
• Data sources: Simulated rainfall events, characterized by nine rainfall intensities and three rainfall patterns.

Main Results

• Rainfall events equivalent to a one-year return period (R5, 42.70 mm total depth) or higher generally yield the most accurate parameters.
• Posterior distribution standard deviations were reduced by 40–60% when using high-intensity rainfalls compared to low-intensity rainfalls.
• Three parameters (impervious area percentage [Imperv], storage depth of impervious area [S-imperv], and Manning’s coefficient of impervious area [N-imperv]) demonstrated consistent accuracy across all rainfall intensities.
• The coefficient of variation for Imperv and S-imperv was below 0.05 across all rainfall intensities.
• Rainfall events with double peaks resulted in more satisfactory calibration compared to single or triple peaks.
• The standard deviation of the Width parameter was reduced from 168.647 meters (single-peak) to 110.789 meters (double-peak).

Contributions

• Establishment of a Bayesian SWMM calibration framework utilizing the DREAM(zs) sampling method.
• Provides valuable insights and guidelines for selecting appropriate rainfall events for SWMM model calibration to improve prediction accuracy for urban non-point pollution control and watershed management.

Funding

Not specified in the provided text.

Citation

@article{Shao2025Bridging,
  author = {Shao, Zhiyu and Wang, Jinsong and Zhang, Xiaoyuan and Du, Jiahui and Yost, Scott A.},
  title = {Bridging Uncertainty in SWMM Model Calibration: A Bayesian Analysis of Optimal Rainfall Selection},
  journal = {Water},
  year = {2025},
  doi = {10.3390/w17233435},
  url = {https://doi.org/10.3390/w17233435}
}