Hydraulically controlled front trapping on a tidal flat

Abstract

[1] We report observations of baroclinic surface fronts in Skagit Bay. The fronts formed repeatedly in very shallow (<0.7 m deep) water along the edge of a tidal channel shortly after the incoming tide flooded the adjacent tidal flat and eventually departed from channel edge as the flood tide progressed. The fronts were surface convergence zones, trapping GPS drifters and other flotsam. Across-front salinity (temperature) jumps were a few practical salinity units (a few degrees) with freshwater on the channel side. Across-front transects of velocity (from mobile and fixed acoustic Doppler current profilers) and density (from mobile and fixed conductivity-temperature-depth profilers) are presented. Composite Froude numbers indicate that supercritical flow over the flats initially held fronts near the channel edge. As flood tide progressed, depth-averaged flow speeds decreased, while internal wave speeds increased, owing to increasing depths and density gradients. Flow over the flats eventually became subcritical, at which point fronts departed from the channel edge. The front then formed the leading edge of a shallow (around 0.3 m thick) freshwater plume which propagated across the flats against the incoming tidal flow. Water was then exchanged between the channel and flats with a channel edge composite Froude number near 0.5, consistent with mixing-induced reduction in channel-flats exchange. The across-channel velocity shear was correlated with the across-channel wind, consistent with a simple eddy viscosity model. Strong winds prevented plume propagation.

A novel drifter designed for use with a mounted Acoustic Doppler Current Profiler in shallow environments

“…The most striking measured vertical flow was an intense downwelling near the channel-edge at times (t ≈ 39 to 45 min) when strong shear was also observed. At this location, previously published observations from fixed instruments show the presence of a surface front, with dense saltwater flowing under a fresher surface layer, and cascading down into the channel from the tidal flats (Mullarney and Henderson 2011). Small wind waves (w = 10-30 mm/s, ~ 1.7 s period) can also be seen (Fig.…”

“…In particular, the problem of phase-wrapping and associated velocity ambiguities (typically encountered by pulse-coherent measurements of moderate or strong velocities in an Eulerian reference frame, Lohrmann et al 1990) was largely mitigated by use of Lagrangian measurements. Deployments over several tidal cycles allowed for identification of regions of bathymetrically steered flow along the edge of a tidal channel (resulting from a channel-edge front, Mullarney and Henderson 2011). ADCPs revealed flow features (such as wave motions, vertical velocities, and depth dependence of horizontal velocity) that would not be resolved by surface drifters alone.…”

“…The surface fronts were commonly observed on days with low to moderate winds. However, on days with high winds (not shown), fronts were not observed and drifters were advected with prevailing winds and surface flows into the salt marsh vegetation toward the East (Mullarney and Henderson 2011).…”

“…3b). As flood tide progressed, fronts departed from the edge of the channel and propagated westwards across the flats as the leading edge of a thin fresh surface plume (Mullarney and Henderson 2011). During deployment 6, drifters were released a second time near the channel edge after the westward spreading of the plume had commenced.…”

Effects of estuarine and fluvial processes on sediment transport over deltaic tidal flats

“…Correlations and skill scores were lower for surface salinity than for bottom salinity. While the across-shore position of the bottom salinity front depended largely on u e (which was well represented by the model), the surface salinity fronts had complex lateral structure and were more sensitive to winds and along-shore currents (Mullarney and Henderson, 2011). While surface salinity fronts in the model were similar in magnitude to the observations, differences in the phasing at the observation points produced large skill errors.…”