Ekman 1993

[Landers Piechotka]

TheMeridional Overturning Circulation (MOC) plays a critical role in global and regional heat and freshwater budgets. Recent studies have suggested the possibility of a southern origin of the anomalous MOC and meridional heat transport (MHT) in the Atlantic, through changes in the transport of warm/salty waters from the Indian Ocean into the South Atlantic basin. This possibility clearly manifests the importance of understanding the South Atlantic MOC (SAMOC). Observations in the South Atlantic have been historically sparse both in space and time compared to the North Atlantic. To enhance our understanding of the MOC and MHT variability in the South Atlantic, a new methodology is recently published to estimate the MOC/MHT by combining sea surface height measurements from satellite altimetry and in situ measurements (Dong et al., 2015).

The goal of this work was to enhance our understanding of the variability of the MOC and MHT in the South Atlantic. In order to achieve this goal, we developed a new methodology that allows to obtain the MOC and MHT time series since 1993 through a combination of the satellite and in situ measurements.

Results obtained from this work show the importance of maintaining sustained observations, such as those from the XBT transect AX18, which are critical to validate altimetry estimates. In addition, measurements from the global observational system are part of the input data of the methodology.

This study represents the first effort to estimate a time series of the MOC in the South Atlantic between 20S and 34.5S. The 20-year time series of the MOC and MHT estimates showed that the dominance of the geostrophic and Ekman components on the interannual variations in the MOC and MHT varies with time and latitude, with the geostrophic component being dominant during 1993-2006 and the Ekman component dominant between 2006 and 2011 at 34.5S (see figure below).

Map of the region together with location of stations and the three sections studied here. Bathymetry from Timmermann et al. (2010). (a) Arrows indicate the Ross gyre and the southern extent of the ACC (after Orsi et al. 1995). Colored circled dots indicate the location of historical CTD stations in the inner basins and outside the shelf break. (b) Location of the three sections. Purple dots indicate CTD stations, and black circled dots indicate the off-shelf stations (occupied during the same cruise) that were used to extrapolate the section edges. Also shown is station 1, which is used as background reference station in section 3.

Sketch of the barotropic velocity, induced Ekman transport, and lifting of isopycnals that lead to slippery Ekman layers. (left) The isopycnals (black thin lines) before lifting and (right) the isopycnals at steady state (i.e., when the baroclinic pressure gradient at the bottom balances the barotropic pressure gradient that drives the main flow).

If it is assumed that physical forcing at the shelf break and over the continental slope exerts a primary control over the cross-shelf transport in the western Antarctic, then some likely processes can be identified for this forcing. Perhaps the simplest is through impingement and flow of the Antarctic Circumpolar Current (ACC) onto the shelf in regions where the slope curves cyclonically into the current path and intrusions of ACC water flow onto the shelf via topographic depressions. The resulting regime, which lacks a shelfbreak frontal system and shows continuous water properties from the deep ocean well onto the shelf, typifies the central coast of the WAP (e.g., Klinck et al. 2004; Klinck and Dinniman 2010). In contrast, a well-defined frontal structure dominates the shelf break and upper slope farther west in the Ross Sea (Gordon et al. 2009; Orsi and Wiederwohl 2009). In this second regime, cross-shelf transport consists primarily of a dense outflow of AABW and a compensating on-shelf flow of warmer offshore waters. Dynamic instabilities either in concentrated shelfbreak currents, as in the Ross Sea or Filchner-Ronne margins (Foldvik et al. 2004) or in association with the ACC in regions where it runs close to or along the slope (Klinck et al. 2004; Beardsley et al. 2004; Dinniman and Klinck 2004) might contribute to cross-slope transport. Similarly, pumping of water across the shelf break might occur in response to migrating weather systems (Thoma et al. 2008; Klinck and Dinniman 2010). The present work focuses on cross-shelf Ekman transport in the frictional boundary layer (e.g., Trowbridge and Lentz 1991; MacCready and Rhines 1991; MacCready and Rhines 1993; Garrett et al. 1993; Whlin and Walin 2001).

The Amundsen Sea falls between the two shelf types characterized by the WAP and Ross Sea slopes. It has no well-defined on-shelf flow of oceanic water from the ACC. Nor is there sufficient water mass modification on the shelf to generate dense outflows and a corresponding sharply defined frontal and current system overlying the shelf break and upper slope. The melting of the Pine Island Glacier requires a flow of oceanic heat from the shelf break to Pine Island Bay. Such a flow of warm ocean water has been observed in one of the deep troughs crosscutting the continental shelf to the Pine Island Bay (Walker et al. 2007) and to the western Amundsen shelf (Whlin et al. 2010). We examine the Ekman mechanism as a potential source for this flow in the frictional transport below along-slope current filaments associated with the ACC.

A presumed along-slope forcing by the ACC can be expected to vary from east to west along the Amundsen Sea in response to proximity of the ACC to the shelf break. Figure 1a shows a sketch of the ACC and the Ross gyre in the Amundsen Sea, as outlined in Orsi et al. (1995). The path of the ACC turns southward as it flows east, the southeastward branch of the Ross Sea gyre and the ACC presumably bifurcating in the offshore region between the Ross and Amundsen Seas. The ACC approaches the continental slope near the central Amundsen Sea and follows it eastward into the Bellingshausen Sea (Orsi et al. 1995). However, measurements of the ACC in the proximity of the Amundsen shelf are sparse, and the bifurcation point is poorly defined.

Here, field results are presented that define slope currents and hydrographic structure along transects in the eastern and western Amundsen Sea (Fig. 1). The dynamics of bottom Ekman transports forced by these slope current conditions are investigated using analytical techniques, and implications for the water mass characteristics in the broad shelf shoreward of the shelf break are discussed.

When Ekman layers become arrested, the associated transport approaches a steady value that is much smaller than the Ekman transport (MacCready and Rhines 1993). However, at the shelf inland of the shelf break, the bottom is horizontal and it is not possible for the Ekman layer to be arrested. Where the bottom is horizontal, the relatively dense water that has been transported up from greater depths can flow south toward the coast (Jacobs et al. 2011). Because this dense water is constantly removed from near the shelf break by the southward flow, the transport in the arrested bottom layer never actually shuts down. It does, however, take place in a layer much thicker than the Ekman layer (MacCready and Rhines 1993), in which the thermal wind balance rather than friction brings the velocity to zero. Because the flow is spread out in a layer that is an order of magnitude thicker than the Ekman layer, the velocities are expected to be comparatively small, consistent with the data (Figs. 3, 4).

The present observations suggest that the bifurcation point between the ACC and the Ross gyre falls somewhere in between 120 and 110W. Eastward flow along the shelf was observed at 110W, and westward flow was observed at 118W. Ekman transport induced by a barotropic eastward current east of 115W would pump warm deep water into the troughs, leading to the eastern shelf basins (Fig. 1), which could explain the presence of such water there. If eastward flow west of 115W would occur, Ekman dynamics would pump water into the western shelf basin, where no warm water originating off shelf was found. The mechanism of arrested Ekman layers in combination with the location of the bifurcation point presents a possible explanation of the observed difference in basin water properties between the eastern and western shelf basins.

3a8082e126