A primitive equation, eddy-resolving numerical model is used to study the inherent time scales of variability in the subtropical ocean, assuming temporally constant surface forcing. Three primary scales arise: mesoscale variability of roughly 50-day period, zonally elongated barotropic bands of 1.1-year period, and basin-scale undulations of approximately 4-year period. The latter are identified as first baroclinic mode Rossby waves, associated with bursts of ventilation at isopycnal outcrops. As a result, the equatorward transport required by Sverdrup theory in the subtropics occurs not as a broad, sluggish drift throughout the interior, but as a succession of more intense flows that slowly propagate westward. The zonally elongated bands agree in characteristics to those predicted from theory of homogeneous turbulence. Eddy energy that is generated by baroclinic instability leaks from baroclinic to barotropic mode and thereafter, due to rotational constraints, seeks low zonal wavenumbers. The gr... Abstract A primitive equation, eddy-resolving numerical model is used to study the inherent time scales of variability in the subtropical ocean, assuming temporally constant surface forcing. Three primary scales arise: mesoscale variability of roughly 50-day period, zonally elongated barotropic bands of 1.1-year period, and basin-scale undulations of approximately 4-year period. The latter are identified as first baroclinic mode Rossby waves, associated with bursts of ventilation at isopycnal outcrops. As a result, the equatorward transport required by Sverdrup theory in the subtropics occurs not as a broad, sluggish drift throughout the interior, but as a succession of more intense flows that slowly propagate westward. The zonally elongated bands agree in characteristics to those predicted from theory of homogeneous turbulence. Eddy energy that is generated by baroclinic instability leaks from baroclinic to barotropic mode and thereafter, due to rotational constraints, seeks low zonal wavenumbers. The gr...