Controlled laboratory experiments have been conducted to study the kinematics of wave-current interactions. The results confirm the conservation of waves under the steady state condition. The data also show that the kinematic effect of the current on waves can be treated as a simple Doppler shift.After the Doppler correction, the kinematics of the waves (either breaking or nonbreaking) follows the linear theory very well. The experiments also confirm the blockage of the waves by currents when the ratio MCo approaches -1/4.

Sea spray within the lowest meter of the atmospheric surface layer was measured with an optical instrument supported on a raft. Sizes of measured droplets appear to follow a two‐segment equilibrium spectrum; the division of two size regions within the production zone (the ejection range of jet drops), however, differs from that outside. Vertical distributions of droplet concentrations vary greatly with the droplet size; critical diameters for the droplet suspension under various wind velocities are determined. The mass concentrations of droplets at various elevations were also obtained. Longitudinally, the droplets are distributed in patches, whose occurrence appears to correlate well with the frequency of dominant waves.

A hot‐film anemometer was used in a wind flume to measure the size and number of water droplets over the air‐water interface. Experiments were conducted both with and without mechanically generated waves and each with both freshwater and saltwater. The effects of wind speed and wave height on the vertical distribution of spray were investigated. The mechanism of spray generation was explored. The results of this study indicate that the vertical distribution of the total horizontal flux of droplets can be expressed by a logarithmic distribution. The drop size distribution (drop size spectrum) at a fixed elevation can be described by a negative power law of drop diameter. The droplet production mechanism by bursting bubbles suggested by Blanchard (1963) is found to account for only a portion of the total production. The effect of spray on the remote sensing of the air‐sea interface is evaluated. The spray is found to affect strongly the measurements obtained from a radiometer.

[1] Observations at the edge of the Loop Current after hurricane Katrina show inertial energy amplified at a depth of approximately 600$700 m. Ray-analysis using the eddy field obtained from a numerical simulation with data assimilation suggests that the amplification is due to inertial motions stalled in a deep cyclone. Citation: Oey,

In a recent paper Lai and Shemdin [1971] report on some measurements of the air turbulence over water waves very similar to some measurements reported earlier by Stewart [1970]. The data in these papers are similar, but Lai and Shemdin have reached very different conclusions from these data. I wish to comment on these differences. ACCURACY OF MEASUREMENTSThe statement of measurement error for the turbulent velocities due to misalignment of the anemometers appears to be incorrect. If the angle of alignment is adjusted during calibration such that the linearized voltages from the anemometers differ by <4%, then, conversely, the measured velocity field will have an error of <4% and not 0.1% as was claimed. Note that these estimates assume that the horizontal velocity u' is of about the same magnitude as the vertical velocity w'. If u' ~ low' [Lai and Shemdin, 1971, Figure 6], the error in w' will be 10 times the error in u' or around 40%.No mention was made of the phase response of the capacitance wave probe. Kinsman [1960] found phase delays of around 0.2 rad in the response of his capacitance probe. MEAN-VELOCITY PROfILEThe mean-velocity profiles reported by Lai and Shemdin [1971] indicate that the value for the friction velocity u, of the boundary layer over mechanically generated waves was larger than its value for a boundary layer in the absence the friction velocity u, of the boundary layer reported by Stewart [1970]. Lai and Shemdin gave no precise reasons for this increase, although their various discussions of the effect implied that it may have been due to the rectification of the velocity fluctuations associated with the Copyright •) 1972 by the American Geophysical Union. nonlinear response of the pitot-static tube (as was reported by Shemdin [1967]). This cannot be the cause. The error due to rectification is most pronounced near the wave surface and becomes negligible approximately three wave amplitudes above the mean water surface. The reported velocity profiles remain semilogarithmic to the lowest levels, and the value of u, is constant from the bottom to the top of the profile. Better insight into the mechanisms involved in the change of u, can be reached by converting the velocity data into the usual nondimensional form used for boundary layer data; i.e., u should be scaled by u, and z scaled by u,/ (kinematic viscosity). The data scaled in this way by Stewart indicate that the change in u, is due to the increased roughness of the surface when waves exist on it. The increased roughness increases the high-frequency part of the wind velocity spectra, as was observed [cf. Lai and Shemdin, 1971, Table 3], and also the turbulent Reynolds stress

Traditionally, investigation of statistical properties of ocean waves has been limited largely to global quantities related to elevation and amplitude such as the power spectral and various probability density functions. Although these properties give valuable information about the wave field, the results cannot be related directly to any portion of the time data from which it was derived. We present a new approach using phase information to view and study the properties of frequency modulation, wave group structures, and wave breaking. We apply the method here to ocean wave time series data and identify a new type of wave group (containing the large “rogue” waves), but the method also has the capability of broad applications in the analysis of time series data in general.

No abstract

Data from a series of deep mooring stations in the Gulf of Mexico (GOM) have been analyzed with the newly developed empirical mode decomposition and Hilbert spectral analysis method, abbreviated as Hilbert–Huang transformation (HHT). The flows in the GOM near the shelf/slope region are treated as a two-layer system, with the 800-m permanent thermocline as the dividing depth. When the data are treated with HHT, motions of different temporal scales are identified. The top layer (depth less than 800 m) is controlled by inertia flow with episodic Loop Current eddies, while the lower layer (depth greater than 800 m) is controlled primarily by the topographic Rossby waves and small-scale cyclonic and anticyclonic eddies. Using a cross-correlation analysis between the appropriate intrinsic mode components from the data, the wavelength, the phase velocity, and the vertical trapping depth for the topographic Rossby waves were determined. Observations are in general agreement with the modeled results by Oey and Lee.