Water Surface Profile Prediction in Compound Channels with Vegetated Floodplains

In the present decade, floods have been among the most dangerous and frequent natural disasters.Most rivers are characterized by compound cross-sections that are usually covered with vegetation. The ability to simulate water surface profiles in vegetated rivers quickly and accurately is crucial in flood forecasting operations. This study aims to introduce a low-cost and practical tool for predicting the water surface profile in compound channels with vegetated floodplains. In particular, the current paper employs the Artificial Neural Network (ANN) and Support Vector Machine (SVM) techniques to develop a model for the prediction of the water surface profile in an experimental channel. For this purpose, two approaches were employed. The first one was based on utilizing non-dimensional data, while the second one used dimensional data.The performances of the prediction methods were determined via a 10-fold cross-validation approach. The comparative results revealed that the SVM algorithm outperformed the ANN and regression models.The performance of the SVM model induced by the dimensional data with a CC of 0.99±0.005 and an MAE of 0.0019±0.0002 was shown to be marginally better than that for the dimensionless data. The sensitivity analysis results also indicated that the relative discharge and relative depth played the most important role in estimating the water surface profile.