资源环境科技发展态势分析平台

An array of three horizontally mounted acoustic Doppler current profilers (ADCPs) deployed in the Fraser River at Mission British Columbia was used to calculate suspended sediment concentration (M), mean grain size (D-g), and geometric standard deviation sigma(g) of the grain size distribution (GSD) using acoustic signal inversion. We examine two different multifrequency methods, an explicit and implicit inversion, by systematically allowing M, D-g, and GSD standard deviation (sigma(g)) to vary in the inversions. This is the first application of these methods to commercially produced ADCPs, and we further develop methods that use both viscous and scattering attenuation. We couple these inversions with bottle samples collected within the ensonified volume that were analyzed for suspended sediment concentration and GSD characteristics. Concentrations ranged from 20 to 350 mg/L with GSD characteristics dominated by silt to fine sand. We estimate the necessary calibration parameters needed for complete acoustic inversions and find that M and D-g are both dependent on the calibration constant, which could be a source of error in multifrequency inversions using uncalibrated ADCPs. Comparisons between the inversion results and samples show that the implicit method tends to perform best at all flows for estimating M while providing realistic estimates of grain size at high flows only. The explicit method performed well at high flows, but poorly at low flows, for M and D-g. Estimates of sigma(g) using both inversion methods indicate frequency dependence. We show that the daily averaged acoustically derived M and D-g capture the hysteretic nature of sediment transport in the Fraser River.