Abstract. This paper presents a study of secondary circulation in a curved stratified channel in northern San Francisco Bay over a 12.5-hour tidal cycle. Secondary currents were strong at times (varying by up to 35 cm/s from top to bottom) but relatively transient, as the balance between centrifugal and lateral baroclinic forcing changed over time. The short travel time around the bend did not allow a steady state balance to develop between centrifugal and lateral baroclinic forcing. During the flood tide the confluence of two streams with different velocities produced a strong lateral gradient in streamwise velocity. As a result, lateral advection was a significant term in the streamwise momentum balance, having the same order of magnitude as the barotropic and baroclinic pressure gradients, and the frictional terms. During the first part of the ebb, secondary currents were induced by lateral baroclinic forcing. The direction of the secondary circulation reversed later in the ebb, as the baroclinic forcing became weaker than the centrifugal acceleration. The gradient Richardson number showed that stratification was stable over most of the tidal cycle, decreasing the importance of friction and allowing secondary currents to persist.

[1] We investigated the driving forces of sediment dynamics at the shoals in South San Francisco Bay. Two stations were deployed along a line perpendicular to a 14 m deep channel, 1000 and 2000 m from the middle of the channel. Station depths were 2.59 and 2.19 m below mean lower low water, respectively. We used acoustic ). Most resuspension events occurred during flood tides that followed wave events during low water. Although wave motions are able to resuspend sediment into the wave boundary layer at low tide, the observed large increases in sediment fluxes are due to the nonlinear interaction of wind waves and the tidal currents.

[1] Observations of density and velocity in a channel in northern San Francisco Bay show that the onset of vertical density stratification during flood tides is controlled by the balance between the cross-channel baroclinic pressure gradient and vertical mixing due to turbulence. Profiles of velocity, salinity, temperature, and suspended sediment concentration were measured in transects across Suisun Cutoff, in northern San Francisco Bay, on two days over the 12.5-hour tidal cycle. During flood tides an axial density front developed between fresher water flowing from the shallows of Grizzly Bay into the northern side of Suisun Cutoff and saltier water flowing up the channel. North of the front, transverse currents were driven by the lateral salinity gradient, with a top-to-bottom velocity difference greater than 30 cm/s. South of the front, the secondary circulation was weak, and along-channel velocities were greater than to the north. The gradient Richardson number shows that stratification was stable north of the front, while the water column was turbulently mixed south of the front. Time-series measurements of velocity and salinity demonstrate that the front develops during each tidal cycle. In estuaries, longitudinal dynamics predict less stratification during flood than ebb tides. These data show that stratification can develop during flood tides due to a lateral baroclinic pressure gradient in estuaries with complex bathymetry.

Lay Abstract The influence of eelgrass (Zostera marina) on near‐bed currents, turbulence, and drag was investigated at three sites in two eelgrass canopies of differing density and at one unvegetated site in the San Juan archipelago of Puget Sound, Washington, USA. The data were compared to the canopy shear‐layer model, which has been shown to describe flow through submerged vegetation in laboratory studies. Although most of the data were consistent with the model, velocity profiles from the site with lowest eelgrass density were not, particularly when currents were strong. The eelgrass attenuated tidal currents by at least 40%. Attenuation decreased with increasing current speed, and was greater than 70% at the site with greatest eelgrass density. The attenuation produces a region of low velocities within the canopy that provides a sheltered environment for zooplankton, benthic invertebrates, juvenile fishes, and microalgae. In contrast, the top of the canopy is a region of strong turbulent mixing, exhibiting large‐scale velocity fluctuations. Friction velocity (an indicator of turbulence intensity) at the canopy top was 1.5–2 times greater than at the sea floor of the unvegetated site. This strong turbulence creates a very different habitat type than within the canopy and promotes exchange of oxygen, nutrients, and organic matter between the canopy and the water above it. The eelgrass canopies generated significantly more roughness and drag than the unvegetated sea floor. The drag coefficient CD for flow over the canopies was 3–8 times greater than at the unvegetated site (0.01–0.023 vs. 2.9 × 10− 3).

[1] We report observations of water surface elevation, currents, and suspended sediment concentration (SSC) from a 10-m deep site on the inner shelf in northern Monterey Bay during the arrival of the 2010 Chile tsunami. Velocity profiles were measured from 3.5 m above the bed (mab) to the surface at 2 min intervals, and from 0.1 to 0.7 mab at 1 Hz. SSC was determined from the acoustic backscatter of the near-bed profiler. The initial tsunami waves were directed cross shore and had a period of approximately 16 min. Maximum wave height was 1.1 m, and maximum current speed was 0.36 m/s. During the strongest onrush, near-bed velocities were clearly influenced by friction and a logarithmic boundary layer developed, extending more than 0.3 mab. We estimated friction velocity and bed shear stress from the logarithmic profiles. The logarithmic structure indicates that the flow can be characterized as quasi-steady at these times. At other phases of the tsunami waves, the magnitude of the acceleration term was significant in the near-bed momentum equation, indicating unsteady flow. The maximum tsunami-induced bed shear stress (0.4 N/m 2 ) exceeded the critical shear stress for the medium-grained sand on the seafloor. Cross-shore sediment flux was enhanced by the tsunami. Oscillations of water surface elevation and currents continued for several days. The oscillations were dominated by resonant frequencies, the most energetic of which was the fundamental longitudinal frequency of Monterey Bay. The maximum current speed (hourly-timescale) in 18 months of observations occurred four hours after the tsunami arrived.

No abstract

, suspended-sediment concentrations (SSC) at 15 and 30 cm above the bed (cmab) increased by 1-2 orders of magnitude, and vertical gradients in SSC were strong enough to produce turbulence-limiting stratification, with gradient Richardson numbers exceeding 0.25. Simultaneously, turbulent stresses (decomposed from wave motions) increased by an order of magnitude. The apparent contradiction of energetic turbulence in the presence of strong stratification was reconciled by considering the turbulent kinetic energy (TKE) budget: in general, dissipation and buoyancy flux were balanced by local shear production, and each of these terms increased during wave events. The classic wave-current boundary layer model represented the observations qualitatively, but not quantitatively since the velocity profile could not be approximated as logarithmic. Rather, the mean shear was elevated by the Stokes drift return flow and wind-generated surface stress, which diffused sediment upward and limited stratification. Our findings highlight a pathway for waves to supply energy to both the production and destruction of turbulence, and demonstrate that in such shallow depths, TKE and SSC can be elevated over more of the water column than predicted by traditional models.

[1] The morphology and evolution of bed forms created by combinations of waves and currents were investigated using an oscillating plate in a 4-m-wide flume. Current speed ranged from 0 to 30 cm/s, maximum oscillatory velocity ranged from 20 to 48 cm/s, oscillation period was 8 s (except for one run with 12 s period), and the median grain size was 0.27 mm. The angle between oscillations and current was 90°, 60°, or 45°. At the end of each run the sand bed was photographed and ripple dimensions were measured. Ripple wavelength was also determined from sonar images collected throughout the runs. Increasing the ratio of current to wave (i.e., oscillatory) velocity decreased ripple height and wavelength, in part because of the increased fluid excursion during the wave period. Increasing the ratio of current to waves, or decreasing the angle between current and waves, increased the three-dimensionality of bed forms. During the runs, ripple wavelength increased by a factor of about 2. The average number of wave periods for evolution of ripple wavelength to 90% of its final value was 184 for two-dimensional ripples starting from a flat bed. Bed form orientations at the end of each run were compared to four potential controlling factors: the directions of waves, current, maximum instantaneous bed shear stress, and maximum gross bed form normal transport (MGBNT). The directions of waves and of MGBNT were equally good predictors of bed form orientations, and were significantly better than the other two factors.