Abstract
The Parsons-Veronis model, based on a two-layer wind-driven ocean, predicts the latitude at which the western boundary current separates from the western boundary. It has been tested on the Gulf Stream using both satellite and in situ observations. The hypothesis attributes the difference in the thermocline depth from the eastern to the western side of the ocean and the corresponding northward geostrophic transport (with closed northern end) to the southward Ekman transport integrated across the basin. Twelve years (1977-88) of satellite sea surface temperature data and wind data [from the Fleet Numerical Oceanography Center (FNOC) wind database] have been used for this study. The satellite-derived Gulf Stream northern edges were used to determine the latitudes of separation (i.e., crossing the 2000-m isobath into deep water). Parsons' model is sensitive to two ``free'' parameters, namely, the reduced gravity and the thermocline depth on the eastern side of the basin. Based on available CTD data and previous current meter studies, these free parameters are selected to establish a representative two-layer model for the midlatitude North Atlantic. When the Ekman drift is integrated over several years, the predicted separation latitude variability agrees with observations with unit slope within 95% confidence limits. The relevant time scale of integration is on the order of 3 years somewhat less than the estimated time for long baroclinic planetary waves to cross the Atlantic. For this limited dataset, little improvement in the prediction is found for a larger number of years of averaging. More detailed and long-term investigation of this hypothesis should be made in the future in context of other western boundary currents.