Using the Software

The HEC-HMS User's Manual (HEC, 2022) provides instructions for developing a hydrologic model using computer program HEC-HMS. That manual describes how to install the program on a computer. It also describes how to use the HEC-HMS graphical user interface (GUI) to create and manage analysis projects; create and manage basin models; create and manage meteorologic models; create and manage HEC-HMS control specifications; create and manage simulation runs; calibrate the models; and review the results. However, using HEC-HMS to gain information required for decision making goes far beyond the mouse-clicking and entering data described in that manual.

Using the Model

To use HEC-HMS to develop information required for planning, designing, operating, permitting, and regulating decision making, the following steps should be taken:

1. Identify the decisions required. This is perhaps the most difficult step in a modeling study: deciding exactly what decisions are to be taken as a consequence of a study. In some cases, this may be obvious. For example, in a flood-damage reduction planning study, the decision to be taken is what measures, if any, to implement to reduce damage in a watershed. In other cases, the decision is not as obvious. However, it is seldom the case that the objective of the study is simply to model the watershed or its channels. Instead, the modeling is a source of information that is to be considered in the decision making.

2. Determine what information is required to make a decision. After the decision that is to be made has been identified, the information required to make that decision must be determined. This subsequently will guide selection and application of the methods used. For example, in a flood-damage reduction study, the hydrologic engineering information required is an annual maximum flow or stage frequency function at an index location. While infiltration plays some role in estimating this frequency function, infiltration information itself is not required for the decision making. Thus the emphasis should be on development of a model that provides peak flow and stage information, rather than on development of a model that represents in detail the spatial distribution of infiltration.

3. Determine the appropriate spatial and temporal extent of information required. HEC-HMS simulation methods are data driven; that is, they are sufficiently flexible to permit application to watersheds of all sizes for analysis of events long and short, solving the model equations with time steps appropriate for the analysis. The user must select and specify the extent and the resolution for the analysis. For example, a watershed that is thousands of square miles can be analyzed by dividing it into sub-watersheds that are hundreds of square miles, by computing runoff from the individual sub-watersheds, and by combining the resulting hydrographs. A time step of six hours might be appropriate for such an application. However, the methods in HEC-HMS can also be used to compute runoff from a two or three square mile urban watershed, using a five-minute time step. Decisions about the watershed extent, about subdividing the watershed, and about the appropriate time step must be made at the onset of a modeling study to ensure that appropriate methods are selected, data gathered, and parameters estimated, given the level of detail required for decision making.

4. Identify methods that can provide the information, identify criteria for selecting one of the methods, and select a method. In some cases, more than one of the alternative methods included in HEC-HMS will provide the information required at the spatial and temporal resolution necessary for wise decision making. For example, to estimate runoff peaks for an urban flooding study, any of the direct runoff methods shown in the Summary of Simulation Methods Included in HEC-HMS will provide the information required. However, the degree of complexity of those methods varies, as does the amount of data required to estimate method parameters. This should be considered when selecting a method. If the necessary data or other resources are not available to calibrate or apply the method, then it should not be selected, regardless of its academic appeal or reported use elsewhere. Furthermore, the assumptions inherent in a method may preclude its usage. For example, backwater conditions eliminate all routing methods in HEC-HMS except Modified Puls, and may even eliminate that method if significant enough.

Finally, as Loague (1985) point out "… Predictive hydrologic modeling is normally carried out on a given catchment using a specific model under the supervision of an individual hydrologist. The usefulness of the results depends in large measure on the talents and experience of the hydrologist …" This must be weighed when selecting a method from amongst the alternatives. For example, if engineers in a USACE district office have significant experience using Snyder's unit hydrograph, this is a logical choice for new watershed runoff analysis, even though the kinematic wave method might provide the same information.

5. Fit model and verify the fit. Each method that is included in HEC-HMS has parameters. The value of each parameter must be specified to fit the model to a particular watershed or channel before the model can be used for estimating runoff or routing hydrographs. Some parameters may be estimated from observation of physical properties of a watershed or channels, while others must be estimated by calibration–trial and error fitting.

6. Collect/develop boundary conditions and initial conditions appropriate for the application. Boundary conditions are the values of the system input - the forces that act on the hydrologic system and cause it to change. The most common boundary condition in HEC-HMS is precipitation; applying this boundary condition causes runoff from a watershed. Another example is the upstream (inflow) flow hydrograph to a channel reach; this is the boundary condition for a routing method. Initial conditions are the known values at which the HEC-HMS equation solvers begin solution of the unsteady flow equations included in the methods. For channel methods, the initial conditions are the initial flows, and for watershed methods, the initial conditions are the initial moisture states in the watershed.

Both initial and boundary conditions must be selected for application of HEC-HMS. This may be a complex, time-consuming task. For example, the boundary condition required for analysis of runoff from a historical storm on a large watershed may be time series of mean areal precipitation (MAP) for subdivision of the watershed. These series would be computed from rainfall observed at gages throughout the watershed, so gage records must be collected, reviewed, reformatted, and processed for each of the gages. Similarly, selection of the initial condition may be a complex task, especially for design applications in which a frequency-based hypothetical storm is used. For example, if the 0.01 AEP flow is required and is to be computed from the 0.01 AEP hypothetical rainfall, the appropriate antecedent moisture condition must be selected. Should a very dry condition be used, or a very wet condition, or some sort of average condition? The choice will certainly have some impact on the model results and hence on the decisions made.

7. Apply the model. Here is where HEC-HMS shines as a tool for analysis. With its graphical user interface and strong data management features, the program is easy to apply, and the results are easy to visualize. As noted earlier, the details of applying the program are presented in the User's Manual (HEC, 2013b).

8. Do a reality check and analyze sensitivity. After HEC-HMS is applied, the results must be checked to confirm that they are reasonable and consistent with what might be expected. For example, the analyst might compare peaks computed for the 0.01 AEP storm from one watershed to peaks computed with the same storm for other similar watersheds. Similarly, the peaks might be compared with peaks computed with other models. For example, if quantiles can be computed with USGS (U.S. Geological Survey) regional regression equations, the results can be compared with the quantiles computed using HEC-HMS and hypothetical rainfall events. If the results are significantly different, and if no good explanation of this difference is possible, then the results from the HEC-HMS model should be viewed with suspicion, and input and assumptions should be reviewed carefully. (As with any computer program, the quality of the output depends on the quality of the input.)

At this point, the sensitivity of results to assumptions should also be analyzed. For example, suppose that the initial and constant loss rate method is used to compute quantiles for flood-damage reduction planning. In that case, the impact of changes to the initial loss should be investigated. If peaks change significantly as a consequence of small changes, and if this in turn leads to significant changes in the design of alternatives, this sensitivity must be acknowledged, and an effort should be made to reduce the uncertainty in this parameter. Similar analyses should be undertaken for other parameters and for initial conditions.

9. Process results to derive required information. In most applications, the results from HEC-HMS must be processed and further analyzed to provide the information required for decision making. For example, if EAD values are required for comparing flood-damage reduction alternatives, the peaks computed for various frequency-based storms must be found in multiple runs of HEC-HMS and must be collected to derive the required flow-frequency function. And if backwater influences the stage associated with the flow, then runs of an open channel flow model may be necessary to develop the necessary stage-frequency function.

ER 1110-2-1464 (USACE, 1994a) provides additional guidance on taking these steps.