This study is concerned with the prediction of the “wind noise” component of ambient noise (AN) in the ocean. It builds on the seminal paper by Felizardo and Melville (1995), in which the authors quantified the correlation between AN and individual wind/wave parameters. Acoustic data are obtained from hydrophones at six diverse locations, and wind/wave parameters are obtained from moored buoys and numerical models. We describe a procedure developed for this study which identifies the correlation of AN with wave parameters, independent of their mutual correlation with wind speed. We then describe paired calibration/prediction experiments, whereby multiple wind/wave parameters are used simultaneously to estimate AN. We find that the improvement from the inclusion of wave parameters is robust but marginal: typically RMSE is reduced by less than 0.3 dB and/or less than 12% of the original RMSE. We interpret the latter outcome as suggesting that wave breaking responds to changes in local winds quickly, relative to, for example, total wave energy, which develops more slowly. This outcome is consistent with a prior study of the dependence of whitecap coverage on wave state, Hwang and Sletten (2008). We discuss this in the context of the time/space response of various wave parameters to wind forcing.
Rogers et al. (Tue,) studied this question.