Expert opinion: uncertainties in hydraulic roughness

Water level predictions in rivers are used for a variety of purposes in water management. For example, designing flood defence measures and evaluating natural rehabilitation in flood plains, cannot be done without water level predictions. However, these water level predictions are uncertain and a major part of this uncertainty is caused by the uncertainty in the roughness coefficient (Van der Klis, 2003; Van Vuren, 2005). Hydraulic roughness in rivers results from (among others): grain roughness, form roughness and vegetation roughness. The roughness coefficient is uncertain because different elements creating the hydraulic roughness are uncertain (e.g. grain size, dune height). To quantify the influence of the uncertain roughness coefficient on water level predictions, we first need a quantification of the uncertainty in the roughness coefficient