Reflections on Rivers, Research, and Rouse

The ways in which rivers adjust their hydraulic roughness to accommodate only loosely correlated changes in their water and sediment discharges are discussed. An assessment of the current status of river engineering leads to the conclusion that the major deficiencies of mathematical models of river flow have been their failure to represent the dependence of friction factor on sediment discharge or concentration, and to include temperature and channel‐curvature effects. It is argued that the data and computational resources now are at hand to permit development of more inclusive analyses of alluvial‐channel flow. A recently developed computer‐based predictor for river‐flow sediment discharge and friction factor, and an analytical model of flow in river bends are outlined. Reflections on Rouse's research and writing conclude the paper.