Computing individual HEC-FIA damages separately is a more complex approach requiring multiple forecast runs and additional processing of consequence results outside of the CAVI. However, this approach provides a few key advantages over the current HEC-FIA Holdouts implementation. First, it allows for computing the without levee damages using HEC-RAS unsteady hydraulics as opposed to the HEC-FIA assumptions. Second, it provides more control over how benefits are apportioned between levees and reservoirs. Third, it allows for computing flood damages reduced for agriculture as well as structures. This approach works best for watersheds with large levees and for large watersheds where reservoirs can impact damage results beyond district boundaries.

In this context, "large levees" describes levees or levee systems with nonnegligible hydraulic impacts to mainstem river stages when comparing with-levee versus without-levee scenarios. In other words, the hydraulic impacts are great enough such that the assumptions used in the HEC-FIA Holdouts compute for the without levee scenario would not be appropriate.

Computing FDR for systems with multiple reservoirs and levees using this approach starts with computing the total damages for the individual scenarios. Those four scenarios are:

  • Regulated: (Observed) Condition with reservoirs and levees in place
  • Unregulated: Condition without reservoirs, but with levees in place
  • Regulated without levee: Condition with reservoirs and without levees
  • Unregulated without levee: Condition with no projects in place


Once the total damages are computed for these scenarios in the CAVI, the process of dividing benefits among reservoirs and levees outside of the CAVI can begin. This is a detailed process involving gathering and computing reservoir allocation percentages using holdouts, computing the damages reduced in leveed areas, and selecting and using an allocation method to allocate portions of levee benefits to reservoirs. This process is discussed in more detail in Interpreting Consequence Results. The remainder of this section describes CAVI setup required beyond a typical CWMS watershed for real-time forecasting.

Forecast Runs

The basic flow of data through the models is shown in Figure 1 and represents the case where the watershed contains at least one reservoir and at least one levee. The model sequence of this forecast run is identical to a standard CWMS real-time forecast run. However, four forecast runs containing different HEC-RAS alternatives for each of the four HEC-RAS scenarios discussed are required. The results of the four HEC-RAS simulations are passed to HEC-FIA to compute total damages for each scenario. At least three new forecast runs need to be created since you can likely reuse an existing observed condition run.  

Forecast Run Schematic - Individual Computes


Figure 2 and Figure 3 show example forecast alternatives for each of the four HEC-RAS scenarios.

 Example Forecast Alternative - With Levee Scenarios

 Example Forecast Alternatives - Without Levee Scenarios

Model Alternatives

Beyond a typical CWMS real-time setup, additional model alternatives are required to produce FDR results. Three additional HEC-RAS alternatives are required to represent the necessary hydraulic scenarios. More details on setting up the alternatives can be found in HEC-RAS Model Setup. Since this approach uses the standard HEC-FIA damage compute, a new HEC-FIA alternative is not required.