Abstract:
Accurate determination of sediment concentration and settling velocity is important to making accurate sediment residence time and transport predictions. However, these parameters can be difficult to measure since optical and acoustic backscatter instruments are inherently sensitive to particle size, shape, and composition – properties that can vary dynamically. In light of this, the present work evaluates the accuracy of traditional and novel sediment sensor calibration and settling velocity methods in a variety of environments and using different instrument configurations in the San Francisco Bay-Delta (SFBD) estuary from 2008 to 2017.

We first determined these sediment parameters using traditional instruments such as the OBS (Optical Backscatter Sensor), ADCP (Acoustic Doppler Current Profiler), and ADV (Acoustic Doppler Velocimeter). We then compared these results against methods that account for sediment flocculation, such as the LISST (Laser In-Situ Scattering and Transmissometry) instrument and two semi-empirical methods that were based on Fractal theory and Rouse theory separated by particle size class. For determining sediment concentration, the OBS had a distinct advantage compared to the LISST when dealing with the effects of a fast vertical profile (> 0.5 m/s), high resuspended sediment concentrations (> 100 mg/L), and schlieren on its optics (N² > 0.025 1/s²). Without these measurement constraints, we observed that the LISST consistently performed the best out of the methods tested. The timing of the peak settling velocity events was also observed to vary widely between the methods.  For example, the LISST and fractal-based settling velocity estimates scaled proportionately with the median floc diameter, while the Rouse-based settling velocity estimates scaled with the vertical sediment concentration gradient.  In the case of the SFBD estuary, we argue that the floc-based methods may be more reliable because the assumptions of Rouse theory broke down in several of the experiments.