Near-bed turbulence and bottom friction during SandyDuck97

[1] Remote acoustic measurements of turbulence intensity profiles are investigated as a function of wave energy and bedstate, from low energy vortex ripples to high energy flat bed. Outside the wave boundary layer, velocity power spectral densities increase with increasing wave energy for all bedstates at frequencies across the wave and turbulence bands up to the Nyquist frequency of the measurements, 8 to 10 Hz. The power spectra of the horizontal and vertical velocity exhibit the -5/3 slope characteristic of inertial subrange turbulence. As the seafloor is approached, the slopes of the vertical velocity spectra in this subrange become progressively less steep, reaching values between -1.2 and -0.6 within the wave boundary layer where the spectral densities are independent of bedstate and incident wave energy. Consistent with laboratory turbulence spectra showing similarly reduced spectral roll-off for the transverse velocity component at microscale Reynolds numbers below 1000 [Sreenivasan, 1996], these field observations indicate that near-bed turbulence generated by irregular waves above a mobile bed is probably anisotropic. Ensemble-averaged vertical turbulence intensity profiles exhibit a peak within the wave boundary layer at heights of O(1 cm) above bottom for all bedstates. The peak is less pronounced for higher energy bedstates. Consistent with the observed spectral convergence within the boundary layer, these peak average turbulence intensities are relatively independent of bedstate, varying by no more than 50% despite a factor of 7 variation in average wave energy. This remarkable observation can be understood from the corresponding decrease in the physical roughness of the bed, associated with the different observed bedstates, as wave energy increases. Estimated wave friction factors are highest for low-energy rippled beds and smallest for flat bed conditions, and within the uncertainty of the measurements, are generally consistent with predictions from the model by Tolman [1994].