| A submersible particle image velocimetry (PIV) system, designed to study the coastal ocean bottom boundary layer, has been built, tested, and deployed. The instrument uses a digital camera and a fiber-optic laser delivery system, and is mounted on a motorized turntable and a hydraulic scissor-jack, allowing traverses of the bottom boundary layer and alignment with the mean current. Details of the design considerations, instrument suite, data acquisition systems, and data analysis are discussed. Data from two deployments, including velocity profiles, turbulence intensities, turbulent spectra, and turbulent production and dissipation estimates, is presented.; In contrast to traditional point sensors, the PIV data yields two-dimensional spatial velocity distributions directly, and allows the computation of four velocity gradients. This data can be used to test the validity of common assumptions such as isotropy, horizontal homogeneity, and Taylor's hypothesis. For instance, the field data suggests that neglecting unresolved components of the production tensor by invoking horizontal homogeneity, may result in significant errors (though this conclusion may be affected by wave contamination). In contrast, the assumption of horizontal homogeneity to estimate normal and shear Reynolds' stresses, introduces only very small errors. Similarly, five different dissipation estimates based on separate sets of assumptions (e.g., isotropic turbulence theory, use of all the measured components of the dissipation tensor, integration of the dissipation spectrum), all yield similar results, though the estimates based on the measured dissipation tensor components are consistently lower than those obtained using assumptions of isotropy.; The turbulent spectra indicate that the flow is anisotropic at all scales. At large, energy-containing scales, the horizontal fluctuations are consistently more energetic than the vertical, even at scales larger than those influenced by waves. At inertial and, surprisingly, even at dissipation scales, the ratio of horizontal to vertical fluctuations is also higher than expected for isotropic turbulence.; Part II describes an image shifting technique, using birefringent and ferro-electric liquid crystals, which can resolve the directional ambiguity of double-pulsed PIV data, and which can be used with non-polarized light sources and fluorescent particles. |
