SWMM Calibration

We are modeling a small (<1 sq. mi.) watershed and have hourly rain gauge data and continuous stream flow rate data. In general, we are pleased with model predictions versus actual measurements, but the model deviates in two areas:

1. Modeled flow rate returns to "base flow" more rapidly than measured flow rate after moderate and large storm events. Peak flow rates match up pretty well. Any suggestions on parameters to adjust?

2. As mentioned above, peak flow rates match up pretty well for moderate and large storms, but grossly overpredict peak flow rates for small storms. We've adjusted depression storage and infiltration parameters, but this hasn't helped significantly. Any suggestions?

Thanks in advance for your help, and thanks UofG for this great service.

Ian,

Bill James has a very nice SWMM Conceptual Workbook that talks about sensitivity analysis and the response of the model to changes in various parameters on an urban catchment. Essentially, small storms are influenced most by PCTZER and % imperviousness. For large storms, changes in slope, width, % imperviousness., maximum infiltration, and minimum infiltration are the key parameters.

Given your situation, I'd suggest examining the following:

1. Try lowering PCTZER and %imperviousness to get a better match on the small storms.
2. If possible, reanalyze the rainfall data to obtain 15 minute or small time increments. This should improve the response of the model to the actual conditions.

Roger W. Lehman, P.E.
Lehmanr@mail.rfweston.com