Influence of combined waves and currents on bedform orientation and evolution

Abstract:

In noncohesive sediments, hydraulic roughness of the seafloor is strongly dependent on bedform morphology, so prediction of bedform morphology is a critical component of sediment transport models. Ripple dimensions are usually predicted from a parameterization of wave energy and bed-sediment grain size. In some coastal environments currents are strong enough to mobilize sediment, but few data are available from either laboratory or field studies to determine the influence of combinations of waves and currents on bedform morphology and evolution. We are investigating the influence of varying combinations of waves and currents on bedform orientation and evolution on the south side of the Strait of Juan de Fuca near the mouth of the Elwha River (water depth 8 m). Median bed-sediment grain size is 130 microns. We measured hydrodynamic forcing and ripple morphology frequently (every 10 to 15 min) to investigate ripple evolution in response to changes in forcing. An imaging sonar captured bedform wavelength (λ) and orientation over 40 m² of the seafloor, and an optical system measured λ, bedform orientation, and height over approximately 0.7 m². Ripples were mostly suborbital (based on the ratio of wave orbital diameter d_o to median grain size), with λ between 10 and 20 cm, and λ decreased with increasing d_o. Ripple heights ranged from 0.5 to 1.5 cm. Bed shear stress was usually dominated by waves, but at times the current shear stress was as great as that of waves. During the two week study period two major shifts in ripple orientation occurred. In the first, the wave direction shifted from 310° to 360° (clockwise from N) over one hour, and ripple orientation shifted from 310° to 350° over 3.25 hours. Following a day of calm conditions northwesterly waves resumed, and the ripple direction shifted back to 310° in 3.25 hours. During periods of relatively low bed shear stress (less than 0.5 N/m², or 3 times the critical shear stress) λ increased gradually, in one case over two days. In contrast, λ decreased rapidly when bed shear stress exceeded 1 N/m² (d_o about 1 m): there were four instances in which λ decreased by 30% over one to two hours. Predictions of bedform morphology can be improved by taking into account the response time of ripples to hydrodynamic forcing, which can be characterized using high frequency data such as these.