Urban drainage models--simplifying uncertainty analysis for practitioners.

Area of Science:

• Environmental Engineering
• Hydrology
• Water Resource Management

Background:

• Increasing awareness of uncertainties in urban drainage modeling necessitates advanced analysis methods.
• Existing uncertainty analysis techniques often have limitations, hindering practical application by engineers.
• Subjectivity and overlooked data uncertainty are common issues in current urban drainage model analysis.

Purpose of the Study:

• To present a modified Monte Carlo-based method for uncertainty analysis in urban drainage models.
• To reduce subjectivity and provide practical, intuitive outcomes for model practitioners.
• To address the common oversight of comparing model uncertainty bands with inherent data uncertainty.

Main Methods:

• A modified Monte Carlo simulation approach is introduced.
• The method compares model uncertainty bands against the uncertainty of observed data points.
• It facilitates intuitive estimation of the optimal number of simulations for uncertainty analysis.

Main Results:

• The developed method provides parameter probability distributions and prediction intervals.
• It offers a less subjective approach compared to traditional methods with cut-off thresholds.
• Demonstrated application on the MOPUS rainfall-runoff model using Melbourne, Australia data.

Conclusions:

• The proposed method enhances the practical application of uncertainty analysis in urban drainage modeling.
• It provides a more objective and intuitive framework for practitioners.
• The approach integrates model and data uncertainty for improved reliability assessments.