Showing the record strengths and growth-rates, recent hurricanes have highlighted needs for improving forecasts of tropical cyclone intensities most sensitive to models of the air-sea interaction. Especially challenging is the nature of sea-spray supposed to strongly affecting the momentum- and energy- air-sea fluxes at strong winds. Even the spray-generation mechanisms in extreme winds remained undetermined. Basing on high-speed video here we identify it as the bag-breakup mode of fragmentation of liquid in gaseous flows known in a different context. This regime is characterized by inflating and consequent bursting of the short-lived objects, bags, comprising sail-like water films surrounded by massive liquid rims then fragmented to giant droplets with sizes exceeding 500 micrometers. From first principles of statistical physics we develop statistical description of these phenomena and show that at extreme winds the bag-breakup is the dominant spray-production mechanism. These findings provide a new basis for understanding and modeling of the air-sea exchange processes at extreme winds.

A turbulent airflow with a centerline velocity of 4 m s−1 above 2.5-Hz mechanically generated gravity waves of different amplitudes has been studied in experiments using the particle image velocimetry (PIV) technique. Direct measurements of the instantaneous flow velocity fields above a curvilinear interface demonstrating flow separation are presented. Because the airflow above the wavy water surface is turbulent and nonstationary, the individual vector fields are conditionally averaged sampled on the phase of the water elevation. The flow patterns of the phase-averaged fields are relatively smooth. Because the averaged flow does not show any strongly nonlinear effects, the quasi-linear approximation can be used. The parameters obtained by the flow averaging are compared with the theoretical results obtained within the theoretical quasi-linear model of a turbulent boundary layer above the wavy water surface. The wave-induced pressure disturbances in the airflow are calculated using the retrieved statistical ensemble of wind flow velocities. The energy flux from the wind to waves and the wind–wave interaction parameter are estimated using the obtained wave-induced pressure disturbances. The estimated values of the wind–wave interaction parameter are in a good agreement with the theory.

Laboratory experiments on studying the structure of the turbulent air boundary layer over waves were carried out at the Wind-Wave Channel of the Institute of Applied Physics, Russian Academy of...

This paper describes the construction of geophysical model function (GMF) for wind speed and surface stress retrieval at high winds from cross‐polarized radar backscatter at the water surface. The starting point is the laboratory experiment designed for the study of X‐band backscattering from water surface. In particular, it was shown that cross‐polarized normalized radar cross section (NRCS) keeps sensitivity to wind friction velocity at high winds alternatively to co‐polarized radar return. Basing on the analysis of the Doppler spectra and simultaneous measurements of parameters of surface waves, we suggest a hypothesis that at high winds the cross‐polarized backscattered signal is formed mainly due to scattering from wave breakers. The hypothesis is supported by the experimentally detected proportionality between the power of the scattered signal at cross polarization and the area swept by wave breakers. These results obtained in the laboratory environment are applied for constructing the similar dependencies for field conditions. Using the phenomenological statistical physics approach, a parameterization of active whitecap coverage fraction dependence on wind friction velocity is suggested. With the use of this parameterization, the empirical function expressing the cross‐polarized NRCS by u_* is derived. Using the surface drag parameterization applicable at strong winds, this dependence is verified on the base of available data sets containing collocated satellite measurements of cross‐polarized C‐band NRCS and ground measurements of wind speed. GMFs for u_* and U₁₀ retrieval are suggested. Analysis shows that taking into account the angular dependence of cross‐polarized radar backscattered power improves the accuracy of wind speed and wind friction velocity retrieval.