River Flow 2012 – Murillo (Ed.) © 2012Taylor & Francis Group, London, ISBN 978-0-415-62129-8 Steep flume experiments with large immobile boulders and wide grain size distribution as encountered in alpine torrents T. Ghilardi & A.J. Schleiss Laboratory of Hydraulic Constructions (LCH), Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland ABSTRACT: Flow conditions and sediment transport capacity in steep mountain rivers are poorly known. One of the main problems is that the presence of macro-roughness elements, such as large relatively immobile boulders, have a strong influence. Experiments carried out in a steep laboratory flume, show that the bed surface occupied by boulders and their protrusion need to be taken into account for the estimation of sediment transport. The tests were performed with a wide mobile grain size distribution and a random position of the large boulders. A clear relationship between the dimensionless distance of boulders and sediment transport capacity is found. The bedload rate can be reduced by 60% when 15% of the bed surface is occupied by boulders, compared to transport rate without boulders. The boulder diameter also has an influence on sediment transport capacity. This is linked to both the surface occupied by the boulders and its average protrusion. 1 INTRODUCTION Flow conditions and sediment transport are well known for lowland rivers. On the contrary, only few studies have been made on steep mountain channels, mainly during the last two decades. Most sediment transport equations, even if developed for high slopes, overpredict sediment flux by several orders of magni- tudes in mountain streams. The reason is that sediment transport equations often don’t take into account the flow resistance induced by the relatively immobile large boulders, which can occupy a large area of the riverbed. Alpine rivers are typically characterized as streams having longitudinal slopes ranging from 0.1% to almost 20% or more (Papanicolaou et al., 2004). These gravel or boulder bed streams constitute an impor- tant part of the total channel length in mountainous regions. Most sediment reaching floodplains are mobi- lized on hillslopes and transit trough high-gradient torrents (Yager et al., 2007). Gravel bed and boulder bed streams are character- ized by a wide grain-size distribution that is composed of finer, more mobile sediment and large, relatively immobile grains or boulders (Rickenmann, 2001; Papanicolaou et al., 2004; Yager et al., 2007). It has been shown that in coarse gravel bed torrents the grain size distribution of the transported bedload approaches that of the bed material only for high flow intensi- ties (Lenzi et al., 1999; Rickenmann, 2001). Ferro (1999) points out that many Sicilian and Calabrian gravel-bed streams have a bimodal bed particle size distribution, characterized by a fine and a coarse com- ponent. Moreover, in theses torrents, the water depth is small compared to the roughness elements. Large relatively immobile boulders can thus be considered as macro-roughness elements. Most sediment transport equations estimate bed- load transport rates based on the difference between critical and total shear stress. Macro-roughness ele- ments induce a certain stress and disrupt the flow by altering the channel roughness (Yager et al., 2007). As Lenzi et al. (2006) underline, if the roughness increases due to the number of boulders, the form drag will also increase. This implies lower shear stresses available at the bed for sediment entrainment. As proposed by many authors, a shear stress par- titioning method is needed to take into account the presence of macro-roughness elements. Different parameters are proposed according to the authors for shear stress and bed resistance equations, but they gen- erally resume to the number of boulders, their cross section, the bed area occupied by them, the distance between boulders and the drag coefficient (Bathurst, 1978; Canovaro et al., 2007;Yager et al., 2007). It is suggested (Yager et al., 2007), that only the part of the shear stress not acting on boulders will induce a sedi- ment transport. Moreover, there is a limited sediment supply because of the bed area occupied by boulders. When commonly used sediment transport formula are adapted in order to take into account only the shear stress acting on mobile sediments and the limited sed- iment availability, the bedload estimation is deeply improved (Ghilardi et al., 2011;Yager et al., 2007). It has been shown that the presence of boulders decreases the sediment transport capacity (Ghilardi et al., 2011; Yager et al., 2007). Boulder dimensionless distance λ/D [-], where λ [m] is the average distance between boulders of diameter D [m], and protrusion P av [m] are good proxies for sediment transport in mountain 407