PMP and PMF for Design by HEC-MetVue and HEC-HMS (2 of 2)

PMP and PMF for Design by HEC-MetVue and HEC-HMS (2 of 2)

This post is to explain how to set up an HMR 52 storm in HEC-HMS by using HEC-MetVue PMS parameters and how to optimize it by setting a objective goal – maximum peak discharge, volume, or stage. Refer to the first post on how to utilize HEC-MetVue to generate an initial PMS from PMP.

As shown in Figure 1, the HEC-HMS meteorology model’s precipitation method needs to set as HMR 52 storm and its parameters can come from the initial PMS by HEC-MetVue – see Figure 11, 12, and 13 of the first post here. The storm center X, Y coordinates are converted from longitude and latitude of PMS using NOAA NGS Coordinate Conversion Tool (NCAT). The Depth-Area-Duration data sets are paired data which also comes from the initial PMS by HEC-MetVue (Figure 2). The HMR 52 storm in HEC-HMS is a 72-hour “nested” storm in which the maximum precipitation depths of various shorter duration storms are embedded (1hr, 6hr, 12h, 24hr, 48hr).

Figure 1
Figure 2

The initial PMS of HEC-MetVue is created aiming at maximizing the precipitation depth over a basin. A PMS with the maximum precipitation depth does not necessarily generate a PMF since the runoff of a basin is not in a linear relationship with the precipitation. For this reason, the HMR 52 storm parameters in Figure 1 will need to be optimized, which can be accomplished by using HEC-HMS built-in Optimization Trial Manager (Figure 3 and 4). The optimization objective is set as peak discharge in Figure 4 and it can be changed to volume, or a stage if a reservoir element is present in the model, depending on the project needs.

Figure 3
Figure 4

All the five parameters of a HMR 52 storm can be optimized as shown in Figure 5, including X, Y coordinates, storm area, orientation, and peak intensity period. From past project experience, if X, Y coordinates are the first two parameters (Parameter 1 and 2) in an optimization trial, the optimization process may go unstable or the objective function is going to have a difficult time to converge; for this reason, X, Y coordinates may better be set as parameter 4 and 5, or they can be left out for optimization since the final result usually is insensitive to X, Y coordinates coming from HEC-MetVue HMR 52 procedures. Also, since the initial parameters entered in HMR 52 storm are from HEC-MetVue and normally they are close to the final optimized values, the parameter can use a relatively small optimization range (Minimum to Maximum) to expedite the process and avoid the optimization trial run becoming unstable.

Figure 5

In an HMR52 storm optimization trial run, the HMR 52 storm parameters will be optimized to certain values (Figure 6) while the objective is achieving its maximum (Figure 7).

Figure 6
Figure 7

The optimization results of the parameters are shown in Figure 8 and the peak discharge comparison is shown in Figure 9 (before and after optimization). For the example model, the peak discharge is only increased by 0.44% after optimization.

Figure 8
Figure 9

2 COMMENTS

comments user
andrew

Hey Jimmy, first of all thanks for posting! This is perfect for what I need to accomplish. However, when I try to run the optimization within HEC-HMS or run an HMR52 storm with my HEC-HMS model, I get a very cryptic warning message and the run/optimization is aborted:

“NOTE 17201: Began computing Optimization analysis “Optimization_PMF” at time 16Sep2021, 14:07:27.
NOTE 20364: Found no parameter problems in meteorologic model “HMR52”.
WARNING 17203: Aborted Optimization analysis “Optimization_PMF” at time 16Sep2021, 14:07:27.”

Have you experienced this before? Any tips on what might be going wrong?

Thanks in advance!

    comments user
    Jimmy

    Not sure the exact reasons. but you may want to try:

    1. change the orders of the parameters you want to optimize;
    2. optimize one parameter at one time to pin down which one gave you the problem;
    3. increase the number of max iterations and/or tolerance;
    4. reduce the range of optimization or the start value of a parameter.

Leave a Reply

Your email address will not be published. Required fields are marked *