Enrico Pandrin, Ph.D. Candidate 37th cycle, University of Trento, DICAM

Direct measurements of bedload in large gravel-bed rivers are challenging due to the high complexity and dynamicity (Figure 1) and therefore it is common to take advantage of indirect methods. These methods aim to provide an estimation of bedload by measuring the morphological changes of the riverbed over time. Although recent developments in topographic survey techniques increased spatial representativeness, the effect of survey frequency has been only partially investigated. Little is known about the temporal scale of scour and deposition processes, its relationship with discharge and its variability with river morphology. The aim of my work is to investigate the role of survey frequency in the provision of reliable morphological change measurements. In particular, I aim at quantifying the bias in volume estimations for different values of formative discharge and related different morphology, from wandering with alternate bars to braided networks and to provide a first estimation of the related temporal scales.

Field surveys are limited to pre- and post-flood measurements, with little chance to increase the survey frequency. Laboratory-based studies can overcome this issue, allowing for a major control on temporal scales, and also ensuring continuous measurements of bedload. Laboratory experiments were performed at the University of Trento, in a 25 m long and 0.6 m wide flume with 0.01 m/m longitudinal slope. The flume was filled with homogeneous 1 mm sediment grain size. A set of 4 constant discharges from 0.7 l/s to 2.0 l/s was used. For each discharge, I performed a series of 9 short runs each followed by a topographic survey. The time between surveys ranged from 15 minutes to 1 hour with the higher value for the lower discharge, having a roughly similar volume of change in each run. Topographic surveys were performed on a regular grid using a spatial discretization of 50x5 mm longitudinally and crosswise, respectively. DEMs of Difference (DoDs) were computed considering all the possible combinations between the 9 surveys. This produced a set of DoDs with different time steps. DoD analysis provided volumes of eroded and deposited sediments, as well as the morphological active width, as the proportion of the channel cross-section over which measurable morphological changes occurred.

Results show a strongly nonlinear relationship between morphological changes and the time between surveys, with a significant and increasing underestimation of the morphological changes for increasing time between surveys. I observed this temporal trend for all four morphologies, with a time scale that depends mainly on the bedload flux and the active width. This time scale ranges from 42 to 115 minutes at the laboratory scale. At the field scale, this could be equivalent to 4-12 hours of constant formative discharge. This means that field surveys separated by longer timesteps are likely to significantly underestimate the occurred morphological changes, with effects on our ability to estimate the bedload flux. In the next month data laboratory outcomes will be validated by comparing data from the field.