In the past, the suspended sediment concentration (SSC) was mainly measured by the optical backscattering device (OBS) and water sample filtration. However, there has been a new development that user is based on the acoustics backscattering (ABS) to measure the SSC in the world. The acoustic instruments have some advantages that the optical ones do not have. For example, acoustic instruments are not effected by high turbidity, biofouling and high viscosity in the water. Acoustic instruments have high spatial and temporal resolutions. And they can immediately indicate the SSC changes than the water sample filtration method. Therefore, in this study we used the multi-frequency acoustics instrument (AQUAscat-1000) to investigate the relations of the suspended sediment size and concentration to the acoustic characteristics. The results are separated into two parts: (1) The calibration process in the flume. (2) The acoustic results in the field experiment. In the first part, we examined the range of preferred bin size with respect to the different frequencies as well as the condition in the flume, and determine that conclude the 20 mm is the best range for our case. In addition, the gain should be used when the maximum SSC is less than 100 mg/l and vice versa. Therefore, in the field case around the river mouth, the signal gain should be turned off due to high concentrations. According to the sensitivity of the backscatter intensity of different frequencies to the suspended particle sizes, multiple frequencies are needed to derive the SSC when the sediment size becomes smaller. The last point in this part is the comparison of the results between the acoustics and optical instruments. When the suspended particles in the water column are transparent, the optical measurement of the SSC is underestimated, but the acoustic method is not. The result of the field experiment in 2009 showed that the AQUAscat-1000 is a better instrument to quantify SSC than the optical instruments. The SSC increase caused by the bottom sediment re-suspension was due to the wave shear. In the 2011 experiment, although the acoustic results overestimated the SSC at some points but they still had higher relation and significance with water sample data than the optical measurements. The SSC increase caused by the bottom sediment re-suspension was due to the current shear.

Cette thèse est une étude de l'applicabilité des profileurs acoustiques Doppler horizontaux (H-ADCP) pour le suivi des flux d'eau et de matières en suspension (MES) dans les rivières. Plus d'un an de données acquises avec des H-ADCP de 300, 600 et 1200 kHz sur quatre sites, sur le Rhône, l'Isère et la Saône, avec des géométries et des conditions de forçage contrastées sont analysées. Les résultats montrent qu'une profondeur de section limitée peut poser problème en raison de la diffusion d'une partie de l'énergie acoustique par la surface libre. De plus, quand l'intensité rétrodiffusée par les particules est trop faible, les mesures de vitesse sont sous-estimées ou plus dispersées par rapport aux mesures de référence. Des relations de vitesse indice sont toutefois établies en fonction de l'intensité et de la concentration afin de corriger les vitesses. La concentration en MES est déterminée à partir de l'atténuation acoustique qui est importante pour des suspensions concentrées de limons (> ~100 mg/L). Les constantes d'atténuation sont obtenues par calage sur des mesures de turbidité~; elles sont proches des valeurs théoriques calculées pour les distributions granulométriques des particules primaires. Les mesures acoustiques de concentration sont en bon accord avec les mesures de référence et reproduisent finement la dynamique temporelle. En outre, l'évolution de la granulométrie est étudiée à partir des mesures multi-fréquences d'atténuation sous l'hypothèse que les distributions granulométriques sont lognormales. Cette étude montre qu'une fois que les conditions limites pour des mesures fiables sont bien établies, le H-ADCP est un outil performant pour le suivi des flux d'eau et de MES dans les rivières, surtout pendant des périodes de fortes concentrations telles que des crues. / The work investigates the feasibility of using horizontal acoustic Doppler current profilers (H-ADCPs) to monitor fluxes of water and suspended sediment in rivers. Year-long data sets acquired with H-ADCPs operating at 300, 600 and 1200 kHz at four sites with varying geometries and flow forcing conditions on the Rhône, Saône and Isère rivers are analyzed. Findings show that limited depth can pose a problem due to scattering of a fraction of the acoustic energy from the air-water interface. A second problem arises when the backscattered intensity from the suspended sediment is too weak; this leads to underestimation or higher variability of the velocity estimates compared to reference values. Nevertheless, index-velocity relationships are established as a function of concentration and intensity in order to correct the velocity measurements. Concentrations of suspended sediment can be determined from the acoustic attenuation, which is substantial for silt-sized particles at concentrations > ~100 mg/L. Attenuation constants obtained by comparison with turbidity data are in good agreement with the theoretical values calculated for the measured grain size distributions of the primary particles. The acoustic measurements of concentration are in good agreement with reference methods in terms of both amplitude and temporal resolution. Grain size is determined from multi-frequency attenuation data, accounting for lognormal grain size distributions. Our findings show that once the limits of accurate velocity measurements are well established for a given site and instrument, the side-looking ADCP can be a valuable tool for monitoring concentration and changes in grain size throughout high concentration events such as floods.