ALBERT

Description: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution February, 1980

Description: The structure of the inertial peak in deep ocean kinetic energy spectra is studied here. Records were obtained from Polymode arrays deployed in the Western North Atlantic Ocean (40°W to 70°W, 15°N to 42°N). The results are interpreted both in terms of local sources and of turning point effects on internal waves generated at lower latitudes. In most of the data, there is a prominent inertial peak slightly above f; however, the peak height above the background continuum varies with depth and geographical environment. Three classes of environment and their corresponding spectra emerge from peak height variations: class 1 is the 1500 m level near the Mid-Atlantic Ridge, with the greatest peak height of 18 db; class 2 includes (a) the upper ocean (depth less than 2000 m), (b) the deep ocean (depth greater than 2000 m) over rough topography, and (c) the deep ocean underneath the Gulf Stream, with intermediate peak height of 11.5 db; class 3 is the deep ocean over smooth topography, with the lowest peak height of 7.5 db. Near f, the horizontal coherence scale is 0(60 km) at depths from 200 m to 600 m, and the vertical coherence scale is O(200 m) just below the main thermocline. A one turning point model is developed to describe inertial waves at mid-latitudes, based on the assumption that inertial waves are randomly generated at lower latitudes (global generation) where their frequency-wavenumber spectrum is given by the model of Garrett and Munk (1972 a, 1975). Using the globally valid wave functions obtained by Munk and Phillips (1968), various frequency spectra near f are calculated numerically. The model yields a prominent inertial peak of 7 db in the horizontal velocity spectrum but no peaks in the temperature spectrum. The model is latitudinally dependent: the frequency shift and bandwidth of the inertial peak decrease with latitude; energy level near f is minimum at about 30° and higher at low and high latitudes. The observations of class 3 can be well-described by the model; a low zonal wavenumber cutoff is required to produce the observed frequency shift of the inertial peak. The differences between the global generation model and the observations of class 1 and class 2 are interpreted as the effects of local sources. A locally forced model is developed based on the latitudinal modal decomposition of a localized source function. Asymptotic eigensolutions of the Laplace's tidal equation are therefore derived and used as a set of expansion functions. The forcing is through a vertical velocity field specified at the top or bottom boundaries of the ocean. For white noise forcing, the horizontal velocity spectrum of the response has an inertial peak which diminishes in the far-field. With the forcing located at either the surface or the bottom, several properties of the class 2 observations can be described qualitatively by a combination of the global and local models. The reflection of inertial waves from a turbulent benthic boundary layer is studied by a slab model of given depth. Frictional effects are confined to the boundary layer and modelled by a quadratic drag law. For given incident waves, reflection coefficients are found to be greater than 0.9 for the long waves which contain most of the energy. This result suggests that energy-containing inertial waves can propagate over great distance as is required by the validity of the model of global generation.

Description: This work was supported by the National Science Foundation through grant OCE 76-80210 and its continuation OEE 78-19833.

Description: The TOPEX/POSEIDON sea level data lead to new opportunities to investigate some theoretical mechanisms suggested to be involved in the El Nino-Southern Oscillation phenomenon in the tropical Pacific ocean. In particular, we are interested in studying the western boundary reflection, a process crucial for the delayed action oscillator theory, by using the TOPEX/POSEIDON data from November 1992 to May 1995. We first projected the sea level data onto Kelvin and first-mode Ross waves. Then we estimated the contribution of wind forcing to these waves by using a single baroclinic mode simple wave model forced by the ERS-1 wind data. Wave propagation was clearly observed with amplitudes well explained by the wind forcing in the ocean interior. Evidence of wave reflection was detected at both the western and eastern boundaries of the tropical Pacific ocean. At the eastern boundary, Kelvin waves were seen to reflect as first-mode Rossby waves during the entire period. The reflection efficiency (in terms of wave amplitude) of the South American coasts was estimated to be 80% of that of an infinite meridional wall. At the western boundary, reflection was observed in April-August 1993, in January-June 1994, and, later, in December 1994 to February 1995. Although the general roles of these reflection events in the variability observed in the equatorial Pacific ocean are not clear, the data suggest that the reflections in January-June 1994 have played a role in the onset of the warm conditions observed in late 1994 to early 1995. Indeed, during the January-June 1994 period, as strong downwelling first-mode Rossby waves reflected into downwelling Kelvin waves, easterly wind and cold sea surface temperature anomalies located near the date line weakened and eventually reversed in June-July 1994. The presence of the warm anomalies near the date line then favored convection and westerly wind anomalies that triggered strong downwelling Kelvin waves propagating throughout the basin simultaneously with the beginning of the 1994-1995 warm conditions.