A simple adaptive model for on-line flood forecasting in large basins is investigated. Alternative configurations of basin elements may be used in the model for the description of basin operation. Two basic versions of the model are considered. The first, denoted as model E, represents a basin as formed by an ensemble of "homogeneous" elements and involves an adaptive estimate of parameters performed using only the discharge observed in real-time at the outlet. The other, model D, is based on an additive adaptive estimate of parameters carried out upstream and considers a basin as a combination of a sub-basin with an interbasin. However, each of these two large geometric elements is split up into smaller elements similar to those of model E. Each model version was tested for a lead time up to 6h by its application to events observed on large Italian basins. It was found that the performance of model E was invariant with basin area and main channel length and that model D was substantially better than model E. The latter conclusion has significant practical value for situations where an appropriate location for an additive flow measuring station can be established.
Evaluation of alternative configurations of basin elements for on-line flood forecasting
CORRADINI, Corrado
1991
Abstract
A simple adaptive model for on-line flood forecasting in large basins is investigated. Alternative configurations of basin elements may be used in the model for the description of basin operation. Two basic versions of the model are considered. The first, denoted as model E, represents a basin as formed by an ensemble of "homogeneous" elements and involves an adaptive estimate of parameters performed using only the discharge observed in real-time at the outlet. The other, model D, is based on an additive adaptive estimate of parameters carried out upstream and considers a basin as a combination of a sub-basin with an interbasin. However, each of these two large geometric elements is split up into smaller elements similar to those of model E. Each model version was tested for a lead time up to 6h by its application to events observed on large Italian basins. It was found that the performance of model E was invariant with basin area and main channel length and that model D was substantially better than model E. The latter conclusion has significant practical value for situations where an appropriate location for an additive flow measuring station can be established.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
