Modeling the wind-driven variability of the South Indian Ocean

Abstract

This article describes the results of numerical experiments carried out with a general circulation ocean model to investigate the effect of the seasonal cycle of the wind forcing on the Agulhas transport. Two cases are described. The first was initialized with temperature and salinity values obtained by horizontally averaging Levitus climatology. The second experiment was designed to isolate the spatial and temporal structure of the barotropic mode. The model, therefore, was initialized with constant values of temperature and salinity. Both experiments were started from rest, forced at their surface with Hellerman and Rosenstein wind stress climatology, and spun up until dynamical equilibrium. According to the experiments there are two distinct modes of variability in the south Indian Ocean. These modes appear to be separated by the topographic ridges that run south of Madagascar. On the western side of the basin there is a dominant mode with a maximum during spring-summer and a minimum during fall-winter. East of Madagascar there is a marked decrease of the circulation in fall and relative maximums during late summer and late winter. The midlatitude time variability, east of 45°E, appears to be dominated by advection and wave propagation. West of 45°E there is dominance by local wind forcing. A comparison between baroclinic and barotropic experiments indicates that although their annual mean structure is markedly different, their monthly anomalies, south of 3O°S, are quite similar. This result, which agrees with previous theoretical and experimental studies, indicates that the seasonal adjustment in the south Indian Ocean is mostly accomplished by the westward propagation of barotropic planetary waves. This propagation is inhibited by the bottom topography of the Madagascar Ridge and Southwest Indian Ridge (∼45°E). These topographic features appear to isolate the Agulhas Current in the western region from the large-scale gyre farther east at seasonal timescales.