The volume transport of the Agulhas Current was measured over a 3-yr period by an array of seven current meter moorings and four current-and pressure-recording inverted echo sounders (CPIES) deployed at 348S. CPIES extended the array farther offshore in order to capture, for the first time, the full Agulhas Current during meander events. Transports derived from CPIES are well correlated with overlapping current meter transports (0.89). The Eulerian mean current is 219 km wide and 3000 m deep, with peak surface speeds of 1.8 m s 21 and a weak northward undercurrent on the continental slope below 1200 m. A new algorithm to capture the western boundary jet transport at each time step T is defined as the poleward transport out to the first maximum of the vertically integrated velocity beyond the half-width of the mean jet. The mean transport of the Agulhas Current jet, so defined, is 284 Sverdrups (Sv; 1 Sv [ 10 6 m 3 s 21 ) with a standard error of 2 Sv. Sampling and instrumental errors are explicitly estimated and amount to an additional 9 Sv. A more traditional estimate, based on net transport integrated to a fixed distance offshore T box , gives a mean transport of 277 6 5 Sv. This transport is 10 Sv greater than an equivalent transport at 328S, corresponding to a latitudinal increase equal to that predicted by Sverdrup dynamics. The time series of T and T box show important differences during solitary meander events and at longer time scales. In terms of an annual cycle, the Agulhas Current appears strongest during austral summer, a similar phase to the Gulf Stream and Kuroshio.

No abstract

[1] This study examines spreading of surface drifter pairs deployed as part of the CLIVAR Mode Water Dynamic Experiment (CLIMODE) project in the Gulf Stream region. The spreading is resolved at hourly resolution and quantified by relative dispersion and finite-scale Lyapunov exponents. At scales from 1-3 km to 300-500 km, the dispersion follows Richardson's law, indicating stirring by eddies comparable in scale to the pair separation distance. At larger scales, the spreading becomes a random walk described by a constant diffusivity. The behavior from 1-3 km to the local deformation radius is inconsistent with the enstrophy cascade of 2-D quasigeostrophic turbulence. To test various hypotheses for this result, drifter pair spreading is examined for pairs that were not launched together, pairs deployed in the eastern subtropical North Atlantic, and CLIMODE pairs subsampled to daily temporal resolution. Our results indicate the presence of significant energy at the submesoscale in the Gulf Stream region which flattens the wave number spectrum and dominates surface stirring at this scale range. Results in the less energetic subtropical eastern Atlantic are more equivocal.

The surface drifting buoys, or drifters, of the Global Drifter Program (GDP) are predominantly tracked by the Argos positioning system, providing drifter locations with O(100 m) errors at nonuniform temporal intervals, with an average interval of 1.2 h since January 2005. This data set is thus a rich and global source of information on high‐frequency and small‐scale oceanic processes, yet is still relatively understudied because of the challenges associated with its large size and sampling characteristics. A methodology is described to produce a new high‐resolution global data set since 2005, consisting of drifter locations and velocities estimated at hourly intervals, along with their respective errors. Locations and velocities are obtained by modeling locally in time trajectories as a first‐order polynomial with coefficients obtained by maximizing a likelihood function. This function is derived by modeling the Argos location errors with t location‐scale probability distribution functions. The methodology is motivated by analyzing 82 drifters tracked contemporaneously by Argos and by the Global Positioning System, where the latter is assumed to provide true locations. A global spectral analysis of the velocity variance from the new data set reveals a sharply defined ridge of energy closely following the inertial frequency as a function of latitude, distinct energy peaks near diurnal and semidiurnal frequencies, as well as higher‐frequency peaks located near tidal harmonics as well as near replicates of the inertial frequency. Compared to the spectra that can be obtained using the standard 6‐hourly GDP product, the new data set contains up to 100% more spectral energy at some latitudes.

Western boundary currents-such as the Agulhas Current in the Indian Ocean-carry heat poleward, moderating Earth's climate and fuelling the mid-latitude storm tracks. They could exacerbate or mitigate warming and extreme weather events in the future, depending on their response to anthropogenic climate change. Climate models show an ongoing poleward expansion and intensification of the global wind systems, most robustly in the Southern Hemisphere, and linear dynamical theory suggests that western boundary currents will intensify and shift poleward as a result. Observational evidence of such changes comes from accelerated warming and air-sea heat flux rates within all western boundary currents, which are two or three times faster than global mean rates. Here we show that, despite these expectations, the Agulhas Current has not intensified since the early 1990s. Instead, we find that it has broadened as a result of more eddy activity. Recent analyses of other western boundary currents-the Kuroshio and East Australia currents-hint at similar trends. These results indicate that intensifying winds may be increasing the eddy kinetic energy of boundary currents, rather than their mean flow. This could act to decrease poleward heat transport and increase cross-frontal exchange of nutrients and pollutants between the coastal ocean and the deep ocean. Sustained in situ measurements are needed to properly understand the role of these current systems in a changing climate.