Estimation of planetary wave parameters from the data of the 1981 ocean acoustic tomography experiment

Using the maximum-likelihood estimation method and minimizationtechniques, quasi-geostrophic wave solutions were fitted to theobservations of the 1981 Ocean Acoustic Tomography Experiment. Theexperiment occupied a 300 km square area centered at 26°N, 70°W, andhad a duration of ~80 days. The data set consisted of acoustictravel-time records, temperature records and CTD profiles, obtainedfrom the acoustic tomographic array, moored temperature sensors andrecorders, and ship surveys, respectively. While the latter twowere conventional spot measurements, the former corresponds tointegral measurements of the temperature (or sound-speed) field.The optimal fit to the data corresponded to 3 waves in the firstbaroclinic mode, evolving under the presence of a westward mean flowwith vertical shear. The flow was estimated to be weak (~2 cm/s),but it changed the wave periods significantly by producing largeDoppler shifts. The waves were dynamically stable to the mean flow,had weak nonlinear interactions with each other and did not form aresonant traid; thus they constituted a fully linear solution.Evidence for the existence of the waves was strongly supportedby the high correlation (~0.9) between the data and the fit, thelarge amount of signal energy resolved (~80 percent), the excellentquality of the wave-parameter estimate (only about 10 percent inerror), and the general agreement between the observations andquasi-geostrophic linear dynamics.

My financial support for the first two years came from theEducation Office at W.H.O.I. My dissertation research wassupported by ONR Grant NOOOl4-82-C0019.

Submitted in partial fulfillment of the requirements for the degree of Doctor of Science at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution August 1985