Theoretical understanding of eastern ocean boundary poleward undercurrents

Abstract

Mean flows along eastern ocean boundaries may be generated by several mechanisms, including tidal rectification [e.g., Georges Bank, Loder (1980)]; melting ice and rivers emptying into the sea [e.g., the northern Gulf of Alaska, Royer (1981)]; forcing by offshore mean density fields (e.g., the Leeuwin Current, Huthnance (1984), and McCreary et al. (1986)]; the nonlinear interaction of forced barotropic shelf waves over a continental margin [Denbo and Allen (1983)]; oscillatory flow over small scale bumps in the shelf bottom topography [Brink (1986), Haidvogel and Brink (1986), and Holloway (1987)]; and mean wind-stress and wind-stress curl [McCreary (1981), Wang (1982), Suginohara (1982), Philander and Yoon (1982), Suginohara and Kitamura (1984), McCreary and Chao (1985), and McCreary et al. (1986)]. We shall concentrate our discussion primarily on the dynamics of poleward-flowing undercurrents on eastern ocean boundaries. Consequently, since mean flows produced by melting ice and rivers are generally surface flows near the coast, this mechanism is not relevant. Tidally rectified flows only exist in special areas where tidal currents and bottom topography slope are large (e.g., Georges Bank), so this mechanism is also not important in the generation of the major poleward-flowing undercurrents. Generation of poleward flow by the nonlinear interaction of forced barotropic shelf waves would also appear not to be a relevant mechanism as the mean flows generated have been estimated to be very small, only ~0.1 cm s−1. Mean flows produced by offshore mean density fields do produce flows of the right strength, but on eastern boundaries these offshore fields drive equatorward rather than poleward undercurrents (McCreary et al., 1986).