Search

Western boundary currents (WBCs) have intensified and become more eddying in recent decades due to the spin-up of the ocean gyres, resulting in warmer open ocean temperatures. However, relatively l...

Key Points: • A regional ocean model is used to examine multidecadal shelf temperature changes on the Agulhas Bank • There are distinct shelf temperature regime changes in 1966 and 1996 • These regime shifts are caused by changes in coastal upwelling linked to large-scale wind variability The Agulhas Bank is an important area for the spawning of small pelagic fish and other species. Here, within a NEMO ocean model, we investigate changes in temperature over the Bank on multidecadal time scales. In agreement with previous observational studies, a shift to colder temperatures is found in 1997. The model also simulates an earlier shift from colder to warmer temperatures in 1966. These shifts are coastally confined and shown, using a climatologically forced model run as a control, to be driven by a north‐south migration in the large‐scale wind belts, rather than by changes in downward heat fluxes or changes in the Agulhas Current itself. The zonal wind changes on the Agulhas Bank show a significant relationship with the Southern Annular Mode, showing some promise for future predictability of cold and warm regimes on the Agulhas Bank. Thus, while the Agulhas Current has been shown in previous work to have a large impact on intra‐annual and interannual temperature variability, this work shows that multidecadal variability in temperature on the shelf is likely to be wind forced.

Large solitary meanders are arguably the dominant mode of variability in the Agulhas Current. Observational studies have shown that these large meanders are associated with strong upwelling velocities and affect the shelf circulation for over 100 days per year. Here 10-year time series from two ocean general circulation models are used to create a composite picture of the Agulhas Current and its interactions with the shelf circulation in meandering and nonmeandering modes. Both models show good agreement with the size, propagation speed, and frequency of observed meanders. These composite meanders are then used to examine the response of shelf waters to the onset of large meanders, with the use of model output enabling the dynamics at depth to be explored. Results show a composite mean warming of up to 3°C of depth-averaged temperature along the shelf edge associated with an intrusion of the current jet onto the shelf driven by an intensification of the flow along the leading edge of large meanders. However, this intensification of flow results in cooling of bottom waters, driving cold events at the shelf break of

International audience; South Indian Ocean eddies (SIDDIES), originating from a high evaporation region in the eastern Indian Ocean, are investigated by tracking individual eddies from satellite data and co-located Argo floats. A subsurface-eddy identification method, based on its steric dynamic height anomaly, is devised to assign Argo profiles to surface eddies (surfSIDDIES) or subsurface eddies (subSIDDIES). These westward-propagating, long-lived features (>3 months) prevail over a preferential latitudinal band, forming a permanent structure linking the eastern to the western Indian Ocean, that we call the 'SIDDIES Corridor'. Key features have been revealed in the mean thermohaline vertical structure of these eddies. Anticyclonic SIDDIES are characterized by positive subsurface salinity anomalies, with subSIDDIES not exhibiting negative surface anomalies, as opposed to surfSIDDIES. Cyclonic subSIDDIES also occur, but their related salinity anomalies are weaker. SubSIDDIES exhibit two cores of different temperature polarities in their surface and subsurface levels. Cyclonic subSIDDIES have their cores at around 150-200 m depth along the 25.4-25.8 kg m-3 potential density layer with anticyclonic subSIDDIES having their cores at 250-300 m along the 26-26.4 kg m-3 density layer. The SIDDIES corridor acts as a zonal pathway for both eddy-types to advect water masses and biogeochemical properties across the basin. This study provides a new insight on heat/salt fluxes, showing that 58% (32%) of the total heat eddy-flux is ascribed to cyclonic (anticyclonic) subSIDDIES, respectively, in the eastern South Indian Ocean. Anticyclonic subSIDDIES have also been found to be the sole high-saline water eddy-conveyor toward the western Indian Ocean.

[1] Tropical cyclone heat potential (TCHP) has been proposed as being important for hurricane and typhoon intensity. Here, a climatology of TCHP is developed for the Southwest Indian Ocean, a basin that experiences on average 11–12 tropical cyclones per year, many of which impact on Mauritius, Reunion and Madagascar, and Mozambique. SODA data and a regional ocean model forced with the GFDL-CORE v.2b reanalysis winds and heat fluxes are used to derive TCHP values during the 1948–2007 period. The results indicate that TCHP increases through the austral summer, peaking in March. Values of TCHP above 40 kJ cm−2, suggested as the minimum needed for tropical cyclone intensification, are still present in the northern Mozambique Channel in May. A time series of TCHP spatially averaged over the Seychelles-Chagos thermocline ridge (SCTR), an important area for tropical cyclones, is presented. The model time series, which agrees well with XBT-based observations (r = 0.82, p = 0.01), shows considerable interannual variability overlaying an upward tendency that matches with an observed increase in severe tropical cyclone days in the Southwest Indian Ocean. Although an increase in severe storms is seen during 1997–2007, the increasing TCHP tendency time series after 1997 coincides with a decrease in total cyclone numbers, a mismatch that is ascribed to increased atmospheric anticyclonicity over the basin. Seasons of increased (decreased) TCHP over the SCTR appear to be associated with dry (wet) conditions over certain areas of southern and East Africa and are linked with changes in zonal wind and vertical motion in the midtroposphere.