On the Adaptive Scheme of a Stage Forecasting Muskingum Model

S. Barbetta, F. Melone, and T. Moramarco (Italy)


Realtime, flood forecasting, lateral inflows, Muskingum method, adaptive scheme.


A simple real-time stage forecasting model recently proposed for equipped river reaches with significant lateral inflows is here modified in order to improve the rising limb future estimate. The model is based on the Muskingum formulation and requires the estimation of five parameters when the downstream rating curve is unknown. The parameter involved in the lateral inflow representation is computed for each time of forecast on the basis of new observations acquired in real-time. The procedure for its estimate is here modified considering the slope of the hydrograph just before the time origin of forecast. This new adaptive procedure is tested through application of the model under real-time forecasting conditions for some flood events occurred in two equipped river reaches of the Upper Tiber basin, in Central Italy. The mean value of stage errors in magnitude and of the coefficient of persistence, both of them computed for all forecasts carried out up to the peak time, shows that the modified model provides more reliable forecast for most of the investigated floods. This result has a significant practical value for flood warning systems based on hydrometric thresholds.

Important Links:

Go Back