Return to Populating Depth Area Duration Dialog
Next, you will run the model simulations using the three different parameter sets for the DeMichele equation. Then you can compare how each parameter set performs. In particular, we'll observe how results differ when using parameters based on a single storm with short duration in comparison to developing parameters from several storms in a catalog with various durations. While storm duration is a variable within the equation, the parameter fit may be better or worse suited to hypothetical storms more or less similar to the observed storms on which the parameters are based.
Set up the Simulation Runs
- Create three Simulation Runs: Hypo All Storms, Hypo Short Storms, and Hypo Storm7.
- For each select the Village_Creek as the Basin Model and Simulation Window as the Control Specification.
- Select the appropriate Meteorologic Model for each.
- Hypo All Storms: Hyp_HMR51_AllStorms
- Hypo ShortStorms: Hyp_HMR51_ShortStorms
- Hypo Storm7: Hyp_HMR51_Storm7
- Compute each of the Simulation Runs.
View and Compare Results
Navigate to the Results tab and open the Graph and/or Summary Table for the NechesRv_Nr_Beaumont sink element for each Simulation Runs you created and the two existing: Hypo Storm 100yr and Hypo Storm48, these simulations are based on User-Specified area reduction curves. Hypo Storm 100yr is based on area reduction from the hypothetical NOAA Atlas 14 100yr 96 hour storm. Hypo Storm47 is a curve based on the 96 hour observed area reduction from a single 96 hour storm, Storm no. 47 from HMR51 that was centered over Elba, AL so it was not included in the Hypo All Storms parameter fitting. Compare the results. Notice that the results for Hyp_HMR51_AllStorms and Hyp_HMR51_Storm7 yield similar results and are in line with historic Storm no. 47, each with a peak flow of approximately 200,000 cfs. The Hypo Storm 100yr simulation produced a peak flow of around 160,000 cfs. The results for Storm57 and Storm65 show a peak flow of around 120,000 cfs. This demonstrates that storms of vastly different duration, even within the same geographic area, may produce inaccurate or varied results in comparison with storms of similar duration from a different geographic area. The function performed well in fitting area reduction factors of a separate storm that was not in the original sampled set of data.
HMR 51 Storm 57 was a 10 hour duration storm and Storm 65 was a 6 hour storm duration, so parameters fit to this singular storm are not taking into account a longer duration. This demonstrates that the storms used to fit the parameters of the DeMichele model should be representative of the storm duration and frequency of interest.
