A study of the velocity structure in a marine boundary layer : instrumentation and observations

The design and operation of a unique flow measuring instrument forbottom boundary layer studies in the marine environment is documented.The effectiveness of the instrument in acquiring data with which modelsof near bottom flows in the ocean can be tested is demonstrated by theresults of a field experiment in Vineyard Sound.The instrument uses four sensors which measure the mean and fluctuatingparts of the three components of the velocity vector at fourheights above the sea bed. The sensors employ the acoustic travel timedifference technique, and are designed to minimize sensor-induced flowdisturbances. BASS, an acronym for Benthic Acoustic Stress Sensor, hasa resolution of .033 cm/sec per least bit, a range of ±62 cm/sec, noiseof .07 cm/sec in 10 sec, and an estimated accuracy of ±.5 cm/sec, referredto an in situ zero point. A complete set of velocity measurementsis made every .750 seconds, each measurement being the vectorcomponent averaged over 15 cm. The data is internally recorded ondigital cassette tape. Eight hours of continuous data can be recorded.BASS was deployed in a tidal flow in Vineyard Sound at a depth of10 m where a time series of u, v, and w velocities at 26 cm, 46 cm,96 cm, and 210 cm above the bottom was recorded. The mean velocity wasdetermined by fitting each 6 hour series with a sixth order polynomialand the deviations from the polynomial, the fluctuating velocity components,were correlated to produce Reynolds stress profiles. Thestress series shows very few negative stress events while the dominantpositive events have an average duration of 5 seconds and exceed30 dynes/cm2.Zero offset was removed from the mean by assuming a log profile atmaximum ebb. Deviations from a log profile developed when the currentdropped below 40% of maximum, i.e., when the flow could no longer be consideredsteady. A break in the Reynolds stress profile at 1 m suggesteda larger length scale than the 1 cm bottom roughness was present in theflow. A value of u* was determined by using the quadratic drag law(u* = 1.56 cm/sec), the log profile method (u* = 1.60 cm/sec), and theeddy correlation method (u* = 1.91 cm/sec). Integral length scalesof 5 m cross-stream, and 2.5 m vertically were identified by correlationcalculations. Two length scales were present in the downstreamdirection, 5 m within 1 meter of the wall and 8 m further from thewall.

Support from the National ScienceFoundation is acknowledged.

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution February 1978