Page - 377 - in Book of Full Papers - Symposium Hydro Engineering

To monitor the efficiency of releases and to avoid environmental degradation, continuous measurements of conductivity, temperature and dissolved oxygen were conducted during the release. Water depth, Suspended Sediment Concentration (SSC) and turbidity were monitored during the releases at two stations. The first station is located immediately downstream of the dam in order to verify the hydrograph and the upstream TSS concentrations. The second station, located in the downstream part of the bypassed reach, serves to verify the diffusion of the hydrograph and of the silt in the bypassed reach. 116 transects spaced at 100-m intervals were surveyed between the Maury dam and the confluence with the Truyère River downstream during the summer preceding the three releases (t0) and in the autumn (t3) to quantify changes before and after the experiments. Within each transect (defined by its bankfull width), water depth, grain size distribution and the fine sediment thickness (silts and sand) were measured. More detailed transects were also surveyed at 5 stations before the first release (t0), after the first (t1), the second (t2) and the third release (t3). The average station length was 250 m and each of them was composed of about 30 transects. These stations were supposed to be more clogged than the average observed on the river. The parameters measured at these cross-sections are the same as those of the previous protocol. At each of these stations, painted plots and scour chains were installed to determine the size of the sediments mobilized during the various releases. At Station 4, coloured sand particles were also injected before the first release and before the third release. This method has been already implemented to track the movement of sands in coastal environments (Ingle, 1966) but few studies have used it in rivers (Grosprêtre, 2011). 400 kg of coloured particles (grain size 2-4mm) were placed on a river cross-section (Fig. 3) before each of two releases. Several sediment samples, of approximately 5 kg, were collected after each of the three releases over several hundred meters until the coloured particles disappeared in the samples. The number of sand particles present in each sample was assessed by image analysis. The sand used in each injection had a different color to improve the two releases comparison. Fig. 3 Photograph of the injection of fluorescent sand tracers on a transect of the Selves River. 377