Predicting channel bed topography in hydraulic falls

We consider inverse methods for predicting the channel bed topography in experiments of hydraulic falls. Nonlinear solutions and weakly nonlinear approximations from Euler-based models are compared to experimental observations. Accurate predictions are obtained for the maximum height of the topography and its constant horizontal level far downstream using the nonlinear method. The weakly nonlinear approximation is shown only to be a good predictor of the maximum height of the topography. The error in the inverse predictions is examined and discussed.