LAUSR

Wood in river plays a significant role in river geomorphology and ecology, by interacting with erosion and sedimentation processes, influencing the nutrients budget and providing physical habitats for a variety of species. However, Large Wood elements (LW, with a trunk diameter > 0.1 cm and a length > 1 m, as defined by Wohl et al. [38] transported during floods can aggravate the expected drawbacks and augment the overall risk at which the urban areas are exposed [7, 8, 28]. LW may cause obstructions along the channel network, mostly in correspondence of bridges or weirs, where it accumulates, clogging the openings and causing a backwater rise upstream. Hydraulic structures can collapse as a result of the increased loads, and unexpected overflowing can occur at the jammed river cross-sections. The formation of LW accumulation is thus a key issue in the hydraulic risk assessment and its mechanism, as well its subsequent modelling, is still object of study [1, 10, 13, 15, 32, 34]. It is highly affected by the interactions between water, sediments and structures, and it is also influenced by the physical features of LW element and river morphology [30]. Due to the potential implications of LW during floods, different types of practical measures have been developed and employed [2, 5] to retain LW upstream of the critical sections (e.g. fins and racks, [26, 33] or to reduce its presence in the river. These measures remain highly linked to the expertise of single practitioners, although recent works have proposed physically based designs of safety structures [25] both in mountain streams and low-land rivers. The need of giving a quantitative prediction of wood transport during floods throughout the river network has led researchers to investigate water-wood interactions both at a laboratory scale [3, 4, 6, 9] and through numerical models [19, 29]. The motion of floating wood on the water surface has been modelled with different strategies [22, 28, 35], mostly based on Eulerian-Lagrangian approaches able to trace wood motion and rotation. As an input, the aforementioned models require the volume and distribution of the wood that can be transported by the flood, and these data are affected by high uncertainty and derived through wood budget at basin scale [16, 37]. In fact, wood entrained in a river can