River flood forecasting with a neural network model

A neural network model was developed to analyze and forecast the behavior o f the river Tagliamento, in Italy, during heavy rain periods. The model mak es use of distributed rainfall information coming from several rain gauges in the mountain district and predicts the water level of the river at the s ection closing the mountain district. The water level at the closing sectio n in the hours preceding the event was used to characterize the behavior of the river system subject to the rainfall perturbation. Model predictions a re very accurate (i.e., mean square error is less than 4%) when the model i s used with a 1-hour time horizon. Increasing the time horizon, thus making the model suitable for flood forecasting, decreases the accuracy of the mo del. A limiting time horizon is found corresponding to the minimum time lag between the water level at the closing section and the rainfall, which is characteristic of each flooding event and depends on the rainfall and on th e state of saturation of the basin. Performance of the model remains satisf actory up to 5 hours. A model of this type using just rainfall and water le vel information does not appear to be capable of predicting beyond this tim e limit.