The mid-latitude ocean's response to time-dependent zonal wind-stress forcing is studied using a reduced-gravity, 1.5-layer, shallow-water
IntroductionThe wind-driven, basin-scale circulation in the mid-latitude upper ocean exhibits variability on different time scales. This variability could be due to changes in the external atmospheric forcing or to the system's intrinsic instability and nonlinearity. The latter explanation was first put forward by Veronis (1963Veronis ( , 1966, but the former has enjoyed greater popularity in the oceanographic community until fairly recently. A systematic application of the methods of dynamical systems theory to the wind-driven circulation problem has yielded several physical mechanisms for the observed lowfrequency variability, on the time scale of several months to several years . This circulation has been studied in the last decade with a hierarchy of models that include rectangular-basin, as well as more realistic configurations. Pedlosky (1996) reviewed work that used time-independent, single-gyre forcing. Single-gyre results of particular interest since then include Berloff's (1997a, 1997b) work that emphasizes the role of basin size in this problem, and Sheremet et al.'s (1997) classification of instabilities for the single gyre. Chang et al. (2001) reviewed double-gyre results and compared the physical mechanisms for intrinsic variability found in this case with those obtained in the single-gyre problem. We only review here, therefore, the results that are most relevant to the present work. These emphasize the double-gyre problem. Jiang et al. (1995; JJG hereafter) studied the double-gyre wind-driven circulation in a mid-latitude rectangular basin on a β-plane using a 1.5-layer, reduced-gravity, shallow-water (SW) model. They used a time-constant zonal-wind profile, symmetric about the basin's mid-latitude axis. Their results showed that when non-linear processes come into play, multiple steady state solutions satisfying identical boundary conditions can arise for sufficiently strong wind-stress forcing. Two stable solutions were found to co-exist over a certain range of parameters. One had a stronger subpolar gyre and the other a stronger subtropical gyre, with overshooting of the western boundary currents to the north and south of the symmetry axis, respectively. Their periodic solutions had periods of several years and several weeks. Following JJG, Speich et al. (1995) analyzed the dependence of the stationary solutions on the SW model's nondimensional parameters such as the amplitude of the forcing, the Ekman number, the Rossby number, the non-dimensional β parameter, and the bottom drag coefficient. Using pseudo-