Date of Award

Spring 1994

Document Type


Degree Name

Doctor of Philosophy (PhD)


Ocean/Earth/Atmos Sciences



Committee Director

Larry P. Atkinson

Committee Member

Shenn-Yu Chao

Committee Member

Gabriel T. Csanady

Committee Member

Chester Grosch

Committee Member

John M. Klinck


Current and temperature data collected along the shelf edge in the South Atlantic Bight were analyzed using a spectral analysis technique. The power spectra of both alongshore currents and temperatures (upon removal of seasonal trends) in the mid- to lower water column suggest a significant energy peak at 28 days. The spatial characteristics of the fluctuations around this period band were determined using the frequency domain empirical orthogonal function (EOF) analysis applied to the concurrent current and temperature records. Consistent results were obtained from the upstream side of the Charleston Bump. Around the 28-day period, temperature seems to have little correlation with currents, suggesting distinct controlling mechanisms over the two variables. Temperature fluctuations in the mid- to lower water column appear to be advected downstream by the mean current. The first and second current EOF modes each represent a southward propagating signal with a wavelength ca. 5000 km and a northward propagating signal with a wavelength ca. 360 km. This modes account for 64.5% and 18.2% of the total normalized variance, respectively. The first mode is probably related to the interaction between topographically induced wave signals and the Gulf Stream current. The wave characteristics of the second mode fit the dispersion relationship sought by Brooks (1978) for similar bottom topographic profile and mean flow conditions. The nature of the second alongshore current mode is therefore likely to be a barotropic shelf wave.

Although the limitation of the current data in the cross-shore direction prohibited calculating the cross-shore shear of the mean flow, the signs of the transfer of energy between the fluctuations and the mean flow were determined. The results were consistent with earlier findings (e.g., Schmitz and Niiler, 1969; Csanady, 1989; Lee, Yoder and Atkinson, 1991) in that the transfer is a two-way process: the fluctuations draw energy from the mean flow and at times also feed back to the mean flow. It seems, though, there are preferable areas where the transfer is a predominantly one-way process.