You should verify that the HEC-HMS model is well calibrated prior to using it for FDR calculations. Model calibration is a necessary to ensure the simulated FDR results adequately match observed data. Since FDR calculations are for past events, leveraging the use of blending can be examined where high-quality observed flow records exist, however blending is not a replacement for calibration. You are expected to calibrate the HEC-HMS model as best as practically possible for the event, because ungagged local flows are needed for the HEC-RAS and HEC-FIA model runs. An uncalibrated HEC-HMS model will propagate errors through the HEC-RAS model and negatively impact FDR computations. Ideally, for a well-calibrated model (simulated flows match observed flows), blending may not be needed. If blending is used, you should to ensure that the observed streamflow data is of good quality and there are not large discrepancies between simulated and observed streamflow at any location.

One of the most important parameters when calculating FDR is the inflow to the reservoir(s). The inflows at each of the reservoirs are the time-series data used to replace of the observed reservoir outflows for the unregulated condition HEC-RAS model. Since FDR is concerned with damages downstream of reservoirs, the upstream locations are important for calibrating the inflows to each respective reservoir. If the computed reservoir hourly inflow time series stored in the CWMS database is highly variable due to the variables used in the computation, it may be best to use the HEC-HMS simulated reservoir inflows to avoid possible instability in the HEC-RAS modeling process for the unregulated simulation. If there is good reservoir inflow data, it should be used instead of the HEC-HMS simulated inflows. Having the appropriate volume into the reservoir is more critical than matching the observed inflows exactly. If there is a high-quality hourly reservoir inflow time series, the HEC-HMS model upstream of the reservoir can be ignored and blending can be leveraged to force the observed inflow into the downstream reaches. This is not true for the downstream reaches of the HEC-HMS model, as those reaches need to be well-calibrated for the local flow inputs to the HEC-RAS model. Figure 1 is an example of a reservoir inflow and downstream stream gage calibrations and review of observed data.

HEC-HMS Calibration of Reservoir Inflow and Downstream Gage

Blending

Blending provides a method for systematically using observed flow to adjust computed streamflow. Blending in HEC-HMS is the process of using observed flow to adjust computed streamflow. If blending in HEC-HMS is enabled, observed flow replaces computed flow and a transition from observed flow to computed flow is applied at the forecast time. The use of blending forces the observed flows into the next routing reach and will reduce the propagation of residual calibration errors downstream throughout the HEC-HMS model. Refer to the HEC-HMS manual for more information on blending.

If blending is used with poor observed data, errors could be introduced into the HEC-HMS model at the calibration points and downstream, as well as throughout the model sequence. Large discrepancies between simulated streamflow and observed flow can cause an abrupt changes in flow over a short period of time if blending is used in HEC-HMS. These abrupt changes in flow could result in instabilities and inaccuracies in the subsequent HEC-ResSim and HEC-RAS model alternatives. Therefore, calibrating the HEC-HMS model to observed flows should be done first, and if blending is used, much care should be taken.