Ecohydraulic investigations studying the effects of hydropeaking on fish are gaining in importance. In most studies, 2D hydrodynamic models use a static definition of roughness coefficients. As a rule, the coefficients have to be adjusted for each discharge, with the general tendency of Manning roughness coefficients to become smaller with increasing discharge. Although this method is sufficient for environmental flow assessment based on stationary flow calculations, the application of such calibrated model for hydropeaking predictions is questionable. Due to high efforts for the measurements of the time-varying water surface, the calibration usually results in a single verification for the base flow.      

In this work, we implement a semi-empirical formulation of the log-law for flows with low relative submergence, and derive a formula for the depth-dependent roughness coefficient. The formula is applied in a case of a 7 km long river reach on the river Lech (southern Germany) to assess the hydropeaking effects on fish species. Hydraulic modeling is performed using the SRH 2D model which implements a linear look-up function for the depth-dependent Manning coefficient in each model element.

The new approach reduces the model calibration efforts to a minimum. The model is able to adjust the roughness coefficients automatically in every model cell depending on the current water stage, reflecting the rapid flow changes due to hydropeaking. Obtained results show good agreement between measured and modelled water surface evolution. The model performance is additionally verified by photogrammetric analysis based on an unmanned aerial vehicle (UAV) aerial imagery.