Breaking Waves on the Ocean Surface

Thesis (Ph.D.)--University of Washington, 2016-08

In the open ocean, breaking waves are a critical mechanism for the transfer of energy, momentum, and mass between the atmosphere and the ocean. Despite much study, fundamental questions about wave breaking, such as what determines whether a wave will break, remain unresolved. Measurements of oceanic breakers, or "whitecaps," are often used to validate the hypotheses derived in simplified theoretical, numerical, or experimental studies. Real-world measurements are also used to improve the parameterizations of wave-breaking in large global models, such as those forecasting climate change. Here, measurements of whitecaps are presented using ship-based cameras, from two experiments in the North Pacific Ocean. First, a method for georectifying the camera imagery is described using the distant horizon, without additional instrumentation. Over the course of the experiment, this algorithm correctly identifies the horizon in 92% of images in which it is visible. In such cases, the calculation of camera pitch and roll is accurate to within 1 degree. The main sources of error in the final georectification are from mislabeled horizons due to clouds, rain, or poor lighting, and from vertical "heave" motions of the camera, which cannot be calculated with the horizon method. This method is used for correcting the imagery from the first experiment, and synchronizing the imagery from the second experiment to an onboard inertial motion package. Next, measurements of the whitecap coverage, W, are shown from both experiments. Although W is often used in models to represent whitecapping, large uncertainty remains in the existing parameterizations. The data show good agreement with recent measurements using the wind speed. Although wave steepness and dissipation are hypothesized to be more robust predictors of W, this is shown to not always be the case. Wave steepness shows comparable success to the wind parameterizations only when using a mean-square slope variable calculated over the equilibrium range waves and normalizing by the wave directional spread. Meanwhile, correlation of W with turbulent dissipation measurements is significantly worse, which may be due to uncertainty in the measurements or bias related to micro-breaking waves. Finally, phase-resolved, three-dimensional, measurements of the whitecaps were made from a new ship-based stereo video system. Comparison with concurrent buoy measurements indicate that the stereo data accurately reproduces the wave statistics, including the frequency spectra. The whitecaps are characterized by transient and spatially localized regions of extreme surface gradients, rather than large crest-to-trough steepnesses. It was found that whitecaps were around 10 times more likely to have extreme slopes, and 50% of the observed extreme surface slopes were in the vicinity of the breaking waves. The maximum whitecap slopes show good agreement with the Stokes 120 degree limiting crest geometry, and the whitecap crest loses much of its maximum steepness shortly after the onset of breaking. The whitecap phase speeds are consistently less than the linear or weakly nonlinear predicted phase speed, which indicate the effect of narrow-band wave groups, despite the broad-band wave spectra.