Knowledge of oceanic winds is vital to the understanding of air/sea interactions, but is difficult to obtain. This thesis investigates a technique for determining wind speeds from acoustic ambient noise data. The results show acoustic ambient noise in the ocean at 4.3, 8.0, and 14.5 kHz to be highly correlated with surface wind speed. Measurement of noise at the sea floor yields estimates of wind speed within ± 1 m/sec when compared with coincident surface observations. Comparisons of ambient noise with coincident surface measured winds has confirmed the frequency dependence of the wind speed noise relation reported by others ( -20 dB/decade). For deployment with a low mean wind speed (4-6 m/sec), separate calibrations are required at 14.5 kHz for wind speeds above and below the speed characterizing white cap formation. The wind noise calibration equations for the Atlantic and the equatorial Pacific agree with each other and with published results within the limits of experimental uncertainty. The effect of the different hydrographies of the two deployment sites seems to be negligible for deployments at similar depths. There is some evidence that a layer of bubbles at the surface will scatter and absorb high frequency noise sig
