Turbulence Measurements in a High Reynolds Number Tidal Channel

Abstract

Measuring oceanic turbulence in a high Reynolds number flow is a challenge for several reasons: strong flows generate high drag on instrument support structures, turbulent fluctuations are intermittent and irregular, and available instrumentation techniques are limited by the spatial and temporal scales they can accurately resolve. Despite these challenges, field measurements are needed to characterize the dynamics of these energetic flows because Reynolds numbers of О(10^8) are not yet achievable in either numerical simulations or laboratory experiments.

This thesis presents the analysis and discussion of turbulence measurements that were acquired in Grand Passage, Nova Scotia, which is a tidal channel where the flow speed reaches 2.5 m/s and the Reynolds number is 8 x 10^7. The data were collected during three separate field campaigns that included the deployment of four bottom-mounted acoustic Doppler current profilers (ADCPs) and an underwater, streamlined buoy “flown” at mid-depth. The data were used to: (1) assess the capabilities and limitations of both instrumentation techniques and analysis methods for turbulence measurements in high-flow environments, (2) characterize the spatial and temporal variability in turbulence and boundary layer parameters, and (3) investigate the validity of existing theoretical and empirical relationships.

The results indicate that speed-bin averaged rates of dissipation, ε, computed from ADCP data, agree to within a factor of two with direct estimates obtained from the shear probes. At all sites, the dissipation rate is log-normally distributed, and spectral and second-order structure function (SF2) methods yield estimates of ε from the ADCP data that agree to within 16%. Doppler noise levels—estimated using a modified SF2 method—are speed-independent and in agreement with those obtained from the velocity spectra.

Spatial variability and ebb/flood asymmetries in both the velocity profiles and the second-order turbulence statistics are attributed—in part—to the upstream bottom roughness. Imbalances in the local rates of production and dissipation are attributed to streamwise advection, and the degree of anisotropy is shown to vary throughout the water column. A modified form of the Kaimal spectrum is shown to predict the ADCP velocity spectra at large scales.