For the first time, the temperature transport of the Agulhas Current is quantified in a time series. Over a 25-month mooring deployment at 34°S, seven tall moorings were instrumented to measure current velocity, temperature, and salinity. Current and pressure-recording inverted echosounders were used to extend geostrophic velocity, temperature, and salinity records to 300 km offshore. In the mean, the current transports 3.8 PW of heat southwards relative to 0°C: -76 Sv at a transport weighted temperature of 12.3°C. A 0.9 PW standard deviation in temperature transport is due to variability in both volume transport and the temperature field. Meandering of the current core dominates variability in the temperature field by warming temperatures offshore and cooling temperatures near the coast. However, meandering has a limited impact on the temperature transport, which varies more closely with a deepening and broadening of the current associated with an inshore isotherm shoaling and an offshore isotherm deepening. Stronger southward temperature transports correspond to a deeper current transporting more volume, yet at a cooler transport weighted temperature. Seasonality is not observed in the temperature transport time series, possibly due to the offsetting effects of cooler temperatures during times of seasonally stronger volume transports. Although volume transport and temperature transport are highly correlated, the large variability in transport weighted temperature means that using volume transport alone to infer temperature transport results in an error which could be as large as 24% of the South Indian Ocean heat transport.

The Agulhas Current, like all western boundary currents, transports salt from the subtropics toward the poles and, on average, acts as a barrier to exchange between the open ocean and the continental seas. Uniquely, the Agulhas jet also feeds a leakage of relatively salty waters from the Indian into the Atlantic Ocean. Despite its significance, the signals and drivers of water mass variability within the Agulhas Current are not well known. To bridge this gap we use 26 months of moored observations to determine how and why salinity – a water mass tracer – varies across the Agulhas Current. We find that salinity variability is driven by both shifting (i.e. changes in location) and pulsing (i.e. changes in strength) of the current. Shifting of the current causes heave and diapycnal mixing of subtropical, central, and intermediate waters. Diapycnal mixing between central and intermediate waters explains most of the variability, creating salinity anomalies between -0.4 and +0.1 psu. Pulsing of the current drives heave and, to a lesser extent, along-isopycnal mixing within the halocline. This cross-stream mixing results in salinity anomalies of up to 0.3 psu. The mean and standard deviation of Agulhas Current volume and salt transports are -76 and 22 Sv and -2650 and 770 Sv psu. Transport weighted salinity has a standard deviation of 0.05 psu. We estimate that O(1013) kg yr−1 of the salt transported southwestward leaks into the fresher Atlantic Ocean. Based on our observations, the variability of the Agulhas Current could alter this salt leakage by an order of magnitude.

The seasonal variability in volume transport of the South Indian Ocean subtropical gyre is characterized for the first time. Only three complete hydrographic crossings of the gyre have been conducted over a 22‐year period, with an upcoming repeat in 2019. Changes to geostrophic transport and thermocline properties imply a strengthening of the gyre from 1987 to 2002. However, some of this strengthening could result from aliasing of seasonal variability. We use data from Argo, satellite altimetry, and an Agulhas Current transport proxy at 34∘S to quantify the seasonal variability of the upper 2,000‐m volume transport. A semiannual cycle is revealed, with peak‐to‐peak amplitude of 6.4 ± 3.1 Sv(1Sv = 106 m3 s−1) and dominated by annual anomalies in quadrature near the eastern and western boundaries. Seasonal aliasing does not account for the observed gyre strengthening.

Indian Ocean meridional heat transport (MHTIO) drives climate and ecosystem impacts, through changes to ocean temperature. Improved understanding of natural variability in tropical and subtropical MHTIO is needed to contextualize observations and future projections. Previous studies suggest that El Niño‐Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) can drive variability in MHTIO. However, it is unclear whether internally generated IOD can drive variability in MHTIO, or if the apparent relationship between IOD and MHTIO arises because both are modulated by ENSO. Here, we use a model experiment which dynamically removes ENSO to determine the role of internally forced IOD on MHTIO. We find that IOD is not linked to anomalies in MHTIO. Nevertheless, internal atmospheric variability drives significant MHTIO variability. There is little evidence for decadal or multidecadal variability in MHTIO, suggesting this may be a region where an anthropogenic trend rises above the level of internal variability sooner.

Mitigation and adaptation strategies for climate change depend on accurate climate projections for the coming decades. While changes in radiative heat fluxes are known to contribute to surface warming, changes to ocean circulation can also impact the rate of surface warming. Previous studies suggest that projected changes to ocean circulation reduce the rate of global warming. However, these studies consider large greenhouse gas forcing scenarios, which induce a significant buoyancy‐driven decline of the Atlantic Meridional Overturning Circulation. Here, we use a climate model to quantify the previously unknown impact of changes to wind‐driven ocean circulation on global surface warming. Wind‐driven ocean circulation changes amplify the externally forced warming rate by 17% from 1979 to 2014. Accurately simulating changes to the atmospheric circulation is key to improving near‐term climate projections.