Statistical models to predict discharge overflow

Area of Science:

• Environmental Engineering
• Hydrology
• Water Resource Management

Background:

• Stormwater overflows pose significant challenges in urban drainage systems.
• Accurate prediction of discharge events is crucial for effective water resource management.

Purpose of the Study:

• To compare the predictive performance of various statistical models against a complex hydrodynamic model for stormwater overflows.
• To evaluate the feasibility of using simpler statistical methods as alternatives to the Storm Water Management Model (SWMM).

Main Methods:

• Probabilistic modeling using logit, probit, Gompertz, and linear discriminant analysis.
• Hydrodynamic modeling employing the Storm Water Management Model (SWMM) with a synthetic rainfall generator (Monte Carlo method).
• Comparison of model outputs with observational data.

Main Results:

• Logit, probit, and Gompertz models yielded predictions comparable to and concordant with SWMM results and observations.
• Linear discriminant analysis underestimated the annual number of stormwater discharges compared to hydrodynamic modeling.
• Statistical models demonstrate potential as viable alternatives to complex hydrodynamic models.

Conclusions:

• Simple statistical models (logit, probit, Gompertz) can effectively predict stormwater overflow operations.
• These models offer a less labor-intensive and complex alternative to traditional hydrodynamic models.
• Successful application of statistical models requires reliable rainfall and overflow level measurements.