4.0 The Hydrologic Ensemble Forecast Service (HEFS) » 4.3 Hydrologic Modeling Uncertainty
Just as the MEFP helps account for uncertainty with meteorological forcing inputs, the Ensemble Post-processes (EnsPost) quantifies the overall uncertainties in the hydrologic response and corrects for biases in streamflow forecasts.
Use the slider to compare the case without the EnsPost applied (first graphic) with the same data after EnsPost has been applied (second graphic).
Question
What are the main differences between the forecasts from before and after EnsPost is applied? (Choose the best response.)
The correct answer is a.
EnsPost increases the spread at all forecast lead times, and this spread reflects the contribution of hydrologic uncertainties to the streamflow forecasting. These include uncertainties originating from the initial conditions, parameters, and equations of the hydrologic models. When the EnsPost is not used and the hydrologic uncertainties are important, the HEFS forecasts may be overly optimistic about the range of possible outcomes and their probabilities (i.e. contain too little spread).
We can create plots of the HEFS Probabilistic Guidance from the ensemble traces, like the ones above. Note how the plots change when EnsPost functionality is used, particularly the higher discharge and stages associated with the 5% and 10% exceedance probabilities. By increasing the ensemble spread, the EnsPost is indicating that the hydrologic modeling introduces uncertainty. It is important to capture this uncertainty or a decision maker may incorrectly assume that a particular outcome, such as flooding, is highly unlikely.
EnsPost may not be included in early implementations of the HEFS. Users of the HEFS should check with their RFC about whether EnsPost is included if it’s not clear from the product label.