Is the Leeuwin Current driven by Pacific heating and winds?

Abstract

Warm west Pacific water can flow through the Indonesian channels to create a basin-scale buoyancy-driven circulation in the Indian Ocean, even in the absence of winds. The driving force for this circulation is the generation of meridional steric height gradients (and associated zonal geostrophic flows) by the cooling of Pacific inflow water towards the latitude-dependent Haney equilibrium temperature. Mass continuity requires that eastern and western boundary currents develop to feed or remove this zonal flow; in particular, a Leeuwin Current-like flow develops at the eastern boundary. To illustrate these ideas, we have run a numerical model of a rectangular “Indian Ocean”, connected via a near-equatorial channel to the “Pacific” — which is treated simply as a reservoir of water with fixed vertical profiles of temperature and salinity. No wind stress curl is applied, so no Sverdrup circulation is produced; an equatorward patch of winds at midlatitudes is introduced in some experiments to allow the possibility of wind-driven upwelling near the eastern boundary. The total mass flux from the Pacific to the Indian Ocean is identically zero in our model, but up to 18.6 × 10 6m 3s −1 flow in each direction between the basins. When our “Pacific” temperature and salinity profiles are as observed in the Indonesian region, cooling to the Haney equilibrium temperature produces a strong eastward flow at midlatitudes, fed by a western boundary current which is in turn fed by inflow from the Pacific. At the eastern boundary a Leeuwin Current develops, with deep mixed layers near Cape Leeuwin; the mixed water feeds a Leeuwin Undercurrent which eventually flows back to the “Pacific” through the western boundary current. A “typical” water flow path in this experiment is thus: out of the Pacific in the top 100m, in a westward zonal jet; poleward along the western boundary current, with upwelling close inshore and heat loss to the atmosphere; broad, slow eastward flow across the basin; some poleward flow (the Leeuwin Current), more heat loss and sinking near the eastern boundary; some equatorward flow (the Leeuwin Undercurrent), to join a broad westward subsurface flow across the basin; and finally, equatorwards subsurface flow along the western boundary, to join an eastward subsurface jet back into the Pacific. The depth-integrated mass transport of the Current-Undercurrent system near Western Australia is near zero, as observed in the Leeuwin Current system. This pattern is hardly affected by the imposition of equatorward winds like those found off Western Australia; i.e. the offshore Ekman drift is overwhelmed by the onshore near-surface geostrophic flow. The timescale for establishment of this flow regime is of order 1000 days, due to the generation of low vertical mode internal Rossby waves at the eastern boundary. This timescale is even larger when the eastern boundary current is better resolved since an alongshore advective timescale also becomes important. When the “Pacific” temperature profiles are replaced by the much colder ones observed in the eastern equatorial Pacific, with no other change, none of these phenomena occur; instead, a typical eastern boundary flow regime is obtained. The equatorward winds generate upwelling and shallow surface mixed layers near the eastern boundary, with equatorward surface flow. A shallow layer is heated at low latitudes, but this effect does not penetrate deep enough to generate significant zonal flows in midlatitudes; the heated water mostly returns to the “Pacific” via the western boundary current and a zonal jet. By contrast with the first experiment, this flow regime takes only about 200 days to become established. Quantitatively the large difference between the specific volume anomaly profiles of the eastern and western equatorial Pacific can be explained (in a double-integrated sense) by the action of zonal winds, which cause a large gradient of pressure (and hence of specific volume anomaly and temperature) along the equatorial Pacific. Thus these winds may be the cause of the large difference between the Leeuwin Current flow regime and other eastern boundary flows — and of the large heat losses to the atmosphere in much of the Indian Ocean.