Cohesive Sediment Modeling In A Shallow Estuary: Model And Environmental Implications Of Sediment Parameter Variation

Numerical models of sediment transport in estuarine systems rely on parameter values that are often poorly constrained and can vary on timescales relevant to model processes. The selection of parameter values can affect the accuracy of model predictions, while environmental variation of these parameters can impact the temporal and spatial ranges of sediment fluxes, erosion, and deposition in the real world. We implemented a numerical model of San Pablo Bay, an embayment within San Francisco Bay, California, for November-December 2014, and compared model outputs to observations of water level, velocity, wave parameters, salinity, and suspended sediment concentration (SSC) in the shallow regions. Idealized model runs show that wind timing relative to the phase of the tides is the strongest control on sediment fluxes and bed erosion. We varied sediment erodibility in the outflow of the Petaluma River; while this causes erosion and deposition to vary strongly through the shallows system, total export from the shallows does not change. Model runs with realistic winds show that wind likely resuspends faster settling particles or allows for more particle flocculation; particle settling velocity controls system-wide sediment accumulation. At the margins of the system, the magnitude of SSC is closely tied to wind direction when winds occur during flood tide, but sediment deposition is less connected: Both bed evolution and SSC need to be considered in the prediction of marsh fate. Spatial patterns of light attenuation due to SSC is strongly tied to assumed settling velocity.Plain Language Summary Numerical models rely on a number of input values, or parameters, to make predictions about sediment movement and accumulation through estuarine systems, yet often the values are poorly known. Selection of parameter values can impact the effectiveness of numerical models; meanwhile, real-world variation in the parameters also impacts sediment transport. We implemented a three-dimensional numerical model in San Pablo Bay, part of San Francisco Bay, California, to predict water level, water velocity, waves, salinity, and suspended sediment concentration to explore these two questions. Our modeling reveals that the timing of winds relative to tidal phase has the strongest impact on spatial patterns of bed erosion and deposition, stronger than wind direction, sediment settling velocity, spring-neap cycle, or spatial variation in bed erosion parameters. When winds are known, settling velocity, coupled with the erosion rate parameter, can impact the spatial distribution of light attenuation, which is important for protecting estuarine systems from phytoplankton blooms. Under our model, a near-shore region will attenuate more light than a near-channel region under low winds only when settling velocity is smaller; that is, model parameter values impact our interpretation of model outputs.