On kinematics and flow velocity prediction in step-pool channels

This paper verifies methods for the prediction of mean flow velocity at the reach scale in mountain streams, investigating the kinematics of a series of two small-scale artificial step-pool sequences and a transitional reach between plane-bed and step-pool under well-controlled hydraulic conditions, and improving the estimation of the energy expenditure between the step crest and the downstream pool. Experimental data were collected using three fish ladder reaches with slopes between 2.6 and 10%. Four types of field measurements were conducted: topographical surveys to extract the thalweg profiles and cross-sectional geometry of reference cross sections; grain size analyses of the bed surface; steady state runs with a given flow rate (0.005–0.234 m3/s), and surveying of the water profile in the most significant cross sections. The following main conclusions were reached: (i) the dominance of spill resistance at the lowest discharge (pool water depth-step height ratios of 0.4) causes primary dimensionless head losses of up to 80%, and these losses progressively decrease to approximately 40% when the water discharge and related pool water depth submerge the upstream step height. A specific predictive equation for the head loss was calibrated and then verified via data from the Rio Cordon. (ii) The verification of literature-sourced equations to predict the reach-averaged flow velocity provided suitable results for several of these equations indicating that the use of a specific step-pool equation does not appear to be crucial to achieving accurate predictions.