Mean Current Structure And Interaction With Bottom Topography

While the large-scale features of the mean surface circulation generally conform to the pattern imparted by wind forcing, this basic pattern is modified or perturbed by interaction of the flow with bottom topography. Sverdrup et al. (1942) pointed out that the configuration of the ACC is strongly influenced by major bathymétrie features.

With the exception of the large array of drifting buoys deployed during FGGE (Garrett, 1980), direct current observations in the Pacific Sector of the Southern Ocean are sparse. The configuration of the flow field has therefore been largely inferred through indirect methods based on the distribution of tracers and the internal density structure (e.g., Reid and Arthur, 1975). In a comprehensive monograph on the South Pacific, Reid (1986) adjusted the flow inferred from tracer patterns with estimated values of bottom currents necessary to satisfy continuity of mass. Only the flow calculated by Reid near western boundaries or in very deep water differs significantly from the patterns that are revealed by tracers alone. We will note such differences later.

A good description of the mean surface circulation of the entire Southern Ocean inferred from the internal density structure was presented by Gordon et al. (1978). Fig. 3.3 is an adaptation of their map (incorporating data from Gordon and Molinelli, 1982) showing the dynamic height of the sea surface relative to 1,000 decibars within the Pacific Sector. The magnitude of the relative geostrophic flow at the sea surface is proportional to the closeness of the contours and is directed parallel to the contours such that higher values are to the left (in the southern hemisphere) of the flow. In this figure, the strong influence of major bathymétrie features on the flow is evident. In the following discussion, observational evidence of such interactions throughout the Pacific Sector is reviewed.

South of Australia, the Southeast Indian Ridge is zonal in orientation and the

Fig. 3.3 Dynamic height of the sea surface relative to the 1,000 decibar level. Units are dynamic metres. Thin line is the 3,000 m isobath. Adapted from Gordon and Molinelli (1982).

most intense flow is along the northern flank of this ridge. Analyses of synoptic sections across the ACC usually reveal that the flow is streaky; that is, concentrated into multiple high velocity cores. These cores are frequently aligned along the flanks of major bathymetric features. Callahan's (1971) analysis of meridional sections along 115°E, 132°E, and 140°E across the ACC reveals an intense jet of eastward flow along the northern flank of the Southeast Indian Ridge and a broader, less-intense, eastward flow south of the ridge. These two eastward cores are separated by a zone of relatively weak flow (eastward or westward) over the ridge crest.

South of Tasmania at 145°E, the Southeast Indian Ridge turns abruptly toward the southeast until it merges with the Antarctic continental rise at approximately 160°E. The intense flow on the north flank of the ridge also turns to the southeast. Downstream of this turn, the trajectories of satellite-tracked drifting buoys (Fig. 3.4) exhibit a persistent wave-like pattern suggestive of a stationary Rossby wave. Such waves are to be expected in eastward flow downstream of large topographic obstructions due to the conservation of potential vorticity (McCartney, 1976).

150'E

leo'E

60 "3

Fig. 3.4. Trajectories of FGGE drifting buoys south of Tasmania and New Zealand. Shaded area is shallower than 3,500 m.

60 "3

150'E

leo'E

Fig. 3.4. Trajectories of FGGE drifting buoys south of Tasmania and New Zealand. Shaded area is shallower than 3,500 m.

Gordon (1972a) reported that the ACC appears to split into three velocity cores as it flows over the Macquarie Ridge at 158°E (Fig. 3.4). Most of the ACC flows through the passage (58°-60°S) south of the southern tip of the Macquarie Ridge, but two other velocity cores coincide with deep gaps in the ridge at 53.5°S and 56°S. Numerous researchers have described the downstream circulation south of New Zealand (e.g., Burling, 1961; Houtman, 1967; Gordon and Bye, 1972; Gordon, 1975; Heath, 1981). The two northern velocity cores appear to merge into a single jet south of the Campbell Plateau. This jet then follows the southeastern flank of the plateau northward to about 50°S where it turns eastward (Fig. 3.3). The southern branch flows northeastward along the northern flank of the Pacific-Antarctic Ridge until it reaches the two major fracture zones (Udintsev and Eltanin) between 150°W and 120°W (Fig. 3.1). Here, the two branches of the ACC converge before crossing the ridge system over the fracture zones.

Corroborating evidence of this inferred flow pattern near the fracture zone is provided by buoy trajectories (Fig. 3.5). Flow approaching from the west is deflected northeastward by the Pacific-Antarctic Ridge. Although some buoys managed to cross the ridge further south, the remaining buoys turned to the southeast near the fracture zones. Two buoys that did not successfully negotiate the Eltanin Fracture Zone were deflected northeastward along the northwest flank of the East Pacific Rise.

Fig. 3.5. Trajectories of FGGE drifting buoys near the Udintsev and Eltanin Fracture Zones. Shaded area is shallower than 3,500 m.

Fig. 3.5. Trajectories of FGGE drifting buoys near the Udintsev and Eltanin Fracture Zones. Shaded area is shallower than 3,500 m.

On the southern flank of the Pacific-Antarctic Ridge, there is a trough in the dynamic topography (Fig. 3.3). The inferred circulation around the trough is clockwise. It is not clear whether the surface flow in this region forms a closed gyre as is known to exist in the Weddell Sea, but topographic constraints would suggest that such a gyre is likely in the deep circulation. By analogy with the Weddell Sea therefore, the flow in this region is referred to as the Ross Sea Gyre. Although the dynamic topography (Fig. 3.3) and iceberg drift trajectories (Tchernia and Jeannin, 1983) suggest that the Ross Sea Gyre is confined to the southwestern sector of the Southeastern Pacific Basin, they may portray only the most intense portion of the flow. Some property distributions (discussed later) suggest that the gyre may extend eastward almost to Drake Passage.

Across the Southeastern Pacific Basin, the dynamic topography (Fig. 3.3) exhibits no persistent sharp gradients, which suggests that the ACC here is characterized by a broad and more-or-less uniform eastward flow until it reaches Drake Passage. This apparent lack of velocity cores may be an artifact of the sparseness of data, or may be related to the rather featureless bathymetry in this region. In any event, conditions appear to be quite different from those within Drake Passage.

Prior to 1975, numerous researchers (e.g., Ostapoff, 1961; Gordon, 1967a; Reid and Nowlin, 1971) presented vertical sections of geostrophic velocity for Drake Passage that exhibited multiple velocity cores. However, the station spacing in these early sections was too wide to resolve the velocity structure adequately. Between 1975 and 1980, this area was intensively examined during field experiments conducted as part of the International Southern Ocean Studies (ISOS) program. One of the major findings of ISOS was that the flow through Drake Passage is almost always organized into three narrow eastward jets separating four wider zones of weaker eastward flow (Nowlin et al., 1977; Whitworth, 1980; Nowlin and Clifford, 1982). These jets are seen to extend to the bottom in meridional setions of potential density, which exhibit three vertically-coherent bands in which isopycnal slope is somewhat steeper than the general downward slope of isopycnals toward the north.

0 0

Post a comment