A Satellite Study of the Biological Significance of
Eddies in the Drake Passage and
A thesis submitted in partial satisfaction of the
requirements for the degree Master of Science
Committee in charge:
B. Gregory Mitchell, Chair
Xi Chen, 1999
All rights reserved
The thesis of Xi Chen is approved:
Signature Page.......................................................................................................... iii
F Geographic latitude
f Coriolis parameter =2W SinF =planetary vorticity
H Water depth
f/H Planetary potential vorticity (PPV)
Figure 2. Southern Ocean krill distribution based on BIOMASS Program results............... 32
Figure 3A. Bathymetry of
Figure 3B. January 1998 and 1999 composite SeaWiFS chl-a for region in 3A................... 33
Figure 3C. Ocean slope variability determined from satellite altimetry for region in 3A 33
Figure 4. Chl-a
in Drake Passage –
Figure 5A. Bathymetry
for Drake Passage –
Figure 5B. 2-year composite SeaWiFS chl-a for same region as 5A.................................... 34
Figure 6A. Sea surface temperature GAC image for April 1990 with drifter track 35
Figure 6B. Sea surface temperature GAC image for March 1996 with drifter track.............. 35
Figure 6C. Sea surface temperature GAC image for July 1996 with drifter track 35
Figure 7A. Satellite sea surface temperature time-series for 3 locations along 58°S 36
Figure 7B. Satellite sea surface temperature for October 1998 for same region as 5A.......... 36
Figure 7C. SeaWiFS surface chl-a for October 1998 for same region as 5A 36
Figure 8A. Bathymetry near 58°S and 54°W 37
Figure 8B. January 1992 sea surface temperature near 58° S and 54°W............................... 37
Figure 9. Conceptual diagrams of ocean current flow over bathymetric features................. 37
I want to thank Greg Mitchell and JL Swan for their assistance in preparing my thesis. I thank Sarah Gille and John Moisan for their helpful discussions, Peter Niiler for surface drifter data and my thesis committee for their helpful comments and encouragement. I want to thank all the people in the Scripps Photobiology Group for their years of support: Scott Cheng, Elizabeth Frame, Mati Kahru, Austin Lee, Hyojik Lee, Todd Loeber, Tiffany Moisan, Jessica Nolan, Rick Reynolds, JL Swan and John Wieland.
NASA JCOSS Fellowship Grant #NGT-542 and NASA Ocean Biology grant # NAG5-6559, which helped support this work.
1992 – 1995 Graduate
1995 - 1997 Assistant
Researcher, Division of Satellite Remote Sensing,
1997-1999 Graduate Student Researcher,
Chen, X., Automatic Tracking and Receiving Microcomputer System for Polar-orbiting Satellites, Marine Forecasts, No. 3 1996, Vol. 13
Chen, X. and B.G. Mitchell. Design Criteria and Preliminary
Applications Algorithms for an Autonomous Profiling Physical-Optical System
(K_SOLO), accepted by AGU 2000 Ocean Science Meeting at
FIELDS OF STUDY
Major Field: Engineering and Ocean Science
Studies in Applied Ocean Science.
Dr. B. Gregory Mitchell and Professor Dariusz Stramski
Institution of Oceanography,
Studies in Satellite Remote Sensing.
Studies in Electrical Engineering.
Professor Changyue Ouyang
Abstract of the Thesis
A Satellite Study of the Biological Significance of
Eddies in the Drake Passage and
Master of Science in Oceanography
B. Gregory Mitchell, Chair
Phytoplankton biomass in pelagic waters of the Southern Ocean is generally
low although there are high concentrations of major inorganic nutrients and
favorable environmental conditions for phytoplankton growth. Synoptic maps of chlorophyll-a (chl-a) concentrations for the entire Southern Ocean collected by the
SeaWiFS ocean color satellite reveal great spatial and temporal variability in
chl-a concentrations during the
austral spring and summer in pelagic Antarctic waters. While most of the pelagic waters south of the
Polar Front (PF) have very low chl-a values, the
The Southern Ocean, defined as all marine waters south of the Subtropical Convergence (STC), is the largest oceanic area with high inorganic macro‑nutrient concentrations and relatively low chl-a (HNLC; see also references in Chisholm and Morel 1991). The paradox of why phytoplankton biomass is generally so low in Antarctic waters during summer has baffled researchers for decades (El‑Sayed 1987). Many researchers have attempted to explain the low phytoplankton biomass as a function of temperature, solar radiation, microelement concentrations, grazing pressure, or effects of deep mixing of the upper water column.
Since the first demonstration by Martin et al. (1990), other researchers have confirmed that iron (Fe) concentrations are low in Antarctic pelagic waters and considerably higher in Antarctic coastal waters. Fe enrichment experiments of natural populations which result in increased chl-a concentrations compared to controls (e.g, de Baar et al. 1990; Buma et al. 1991; Helbling et al. 1991; Martin et al. 1990) support the hypothesis that Fe limits biomass in Antarctic pelagic waters when there is sufficient light; in winter, insufficient light leads to low biomass. These experiments have demonstrated that with sufficient Fe and light, phytoplankton biomass can increase considerably above ambient concentrations, and macronutrients can be completely consumed within relatively short periods of time (days-weeks). Surface light
is abundant for months, so the
inability of phytoplankton to consume the available macronutrients during the
summer is generally considered to be caused by limitation of Fe. Still, there is large variability in surface
chl-a within pelagic waters of the Southern Ocean. If Fe is the main nutrient limiting biomass
accumulation, then its distribution and rates of supply to the euphotic zone
must also be variable within this important region that comprises approximately
20% of the global oceans and has a key role in global biogeochemical
processes. This research focuses on
physical mechanisms that may contribute to variable rates of Fe supply in the
Drake Passage region and the persistence of the large gradient in chl-a which is low to the west and high to the east of
SeaWiFS chl-a distributions
Ship surveys consistently have indicated that pelagic Antarctic waters have low chl-a values, and that waters in the PFZ and in the Sub-Antarctic zone show higher chl-a values (e.g., Hays et al. 1984; Froneman et al. 1995). These data are consistent with the satellite-derived maps for chl-a in the Southern Ocean (Figure 1A), which shows elevated chl-a concentrations to the north of the PF. While chl-a values for the Southern Ocean are generally low, there is much variability during January in surface chl-a concentrations in Antarctic pelagic waters south of the PF (Figure 1A). Since January is the period of maximal solar radiation and ocean thermal stratification, this is the season when one might expect maximum phytoplankton biomass (if the system were light limited). Conspicuous differences evident in Figure 1 are: (i) Persistent regions of blue water with surface chl-a concentrations less than 0.2 mg m-3 during mid‑summer for vast regions of the Antarctic Circumpolar Current (ACC) south of the Antarctic Polar Front (PF). The largest contiguous blue water zone (BWZ) extends from east of the Ross Sea to north of the South Shetland Islands and into the Drake Passage.
The seasonal changes in SeaWiFS-derived monthly mean surface chl-a concentrations in five regions at 60°S indicated in Figure 1A are illustrated in Figure 1B. All regions were south of the PF except region #5, which was in the Polar Front Zone but on the north side of the PF. Data for region #1 (south of the PF and located in the BWZ) and region #3 show that certain regions of the ACC simply do not develop significant blooms at any time of the year. Values of chl-a in October in the various BWZ regions are equivalent to other regions, but rather than increasing to a mid‑summer maximum, their chl-a decreases after November and remains low through the summer and autumn. By contrast chl-a peaks in December for regions #2, #4 and #5. In January, BWZ waters have surface chlorophyll values of approximately 0.1 mg m-3. The BWZ regions are evident in the 6-year mean CZCS data (Sullivan et al. 1993). OCTS had similar temporal coverage as SeaWiFS and the BWZ regions are evident in that data during the 1996-1997 season (data not shown). The data in Figure 1B and OCTS data (not shown) indicate that the BWZ regions persist through the austral spring‑summer season, and recur each year in the same regions. While these satellite maps only provide information on chl-a concentrations of the upper mixed layer (UML), the decline in surface pigments in BWZs from early spring through summer, rather than a mid-summer peak observed in other regions, implies that surface biomass for BWZs is more severely limited by some essential nutrient(s), since light, temperature and upper ocean stratification increase from spring to summer. By January the difference between surface chl-a in region #1 and #2 represents the extremes observed in the pelagic waters of the Southern Ocean south of the PF. The high phytoplankton biomass in region #2 - the western Scotia Sea - must be supported by elevated fluxes of Fe which is now accepted to be the limiting nutrient for the high nutrient, low chlorophyll (HNLC) regions of the Southern Ocean.
Because the mean flow is from west to east in the Drake Passage, there must be mechanisms that modify the extremely blue water that enters the central Drake Passage, and transforms it into high chl-a water a very short distance to the east. While there is general agreement that Fe limits biomass of phytoplankton in the Southern Ocean, there is relatively little information about why there is so much variability in the distribution of chl-a in the region. This study focuses on physical mechanisms that may contribute to the regional variability in chl-a in the pelagic zones of Drake Passage (areas #1, #2 and #5 in Figures 1A & 1B). This area contains regions with the minimum and maximum annual cycle of chl-a observed for open waters of the ACC south of the STC. As shown in Figure 2, the eastern part of this study region, in particular the zone northeast of Elephant Island, is among the regions with the highest krill biomass in the Southern Ocean. The transformation of BWZ water with extraordinarily low chl-a in the west of the study region into the high chl-a water to the east which can sustain high krill biomass has not been explained. This study focuses on physical mechanisms that may contribute to this extreme gradient in biological response.
Oceanographic characteristics of the region
The Antarctic Circumpolar Current (ACC) is associated with 3 fronts, the Sub Antarctic Front (SAF) on the north, Polar Front (PF) in the middle and Southern ACC Front (SACCF) in the south. There are large surface temperature gradients across these fronts, with warm water in the north and cold water in the south.
There are several water masses in the study region that must be considered in a study of physical, chemical and biological oceanography of the Drake Passage and Western Scotia Sea. Two dominant water masses south of the Polar Front (PF) are the Antarctic Surface Water (AASW) to about 200m, and Circumpolar Deep Water (CDW). Modifications of both water masses at the northern and southern boundaries of the BWZ may be important for understanding the variability of chl-a distribution. The frontal zones on the north and south of the BWZ are the Polar Front (PF) and the Continental Water Boundary (CWB) (Nowlin and Clifford 1982). North of the PF is the Sub-Antarctic Front (SAF). The SAF and the CWB mark the northern and southern boundaries of the east flowing ACC. The CWB marks the boundary of the east flowing Antarctic slope current (Whitworth et al. 1998), which is the East Wind Drift of earlier researchers. The Antarctic slope current is very narrow or non-existent at the northern tip of the Antarctic Peninsula. The eastern boundary of the BWZ is approximately the western boundary of the Weddell-Scotia Confluence (WSC) (Patterson and Sievers 1980, Foster and Middleton 1984) where waters from the Weddell Sea meet those from the eastern Pacific.
In the Pacific Ocean, CDW is derived from interior waters of the Pacific and Indian Oceans and is designated Upper Circumpolar Deep Water (UCDW). In the Atlantic sector this ubiquitous water mass is derived from North Atlantic Deep Water (NADW) and is referred to as Lower Circumpolar Deep Water (LCDW). The sector of the Antarctic Ocean of concern here (the South American Sector), includes both types of CDW, as well as modified forms derived from the Weddell Sea and the Antarctic Peninsula (Whitworth et al. 1998; Hoffmann and Klinck 1998). In general, as CDW upwells towards the surface near the continent, an interesting characteristic of the summer water column appears. The Antarctic Surface water (AASW) contains a strong subsurface temperature minimum (Tmin) south of the PF almost everywhere over deep Antarctic circumpolar waters. Mosby (1934) called this Tmin layer "winter water". As surface waters warm by insolation and deepen by wind mixing, a remnant of the deep-mixed winter layer remains intact near 100 m, with temperatures often < -1.0°C. This Tmin layer isolates CDW which will have re-mineralized iron from the euphotic zone of the BWZ. Of particular interest for this study are the possible mechanisms that allow CDW water to be mixed to the surface and permit cross-shelf transport from both the Antarctic Peninsula and the South American shelves into the central region of the eastern Drake Passage as the current flows into the Scotia Sea.
Despite significant interest in understanding the role of Fe in moderating chl-a concentrations and primary production in the Southern Ocean there are few studies documenting dissolved Fe concentrations. Martin et al. (1990) reported values of Fe ranging from 0.1-0.16 nM for 3 depths from the surface to 110 m in the eastern part of the BWZ (see station position in Figure 3B). The value within the 110 m deep temperature minimum layer (discussed above) was 0.1 nM Fe. Values north of the polar front were similar, while values in Gerlache Strait were much higher. Along 6°W, Loscher et al. (1997) reported values of Fe in surface waters south of 50°S that were typically about 0.5 nM, with generally higher values at 100 m. Timmermans et al. (1998) reported Fe concentrations from 0.2-0.5 nM in surface waters near 90°W, including stations that would be in the BWZ south of the APF as well as the region north of the APF.
Concentrations at only a few stations and depths as reported in the literature are not sufficient to understand the role of Fe in controlling seasonal biogeochemical dynamics within the ACC system. Resupply of Fe to surface waters can occur via three major routes: (i) from below, through the diffusion or upwelling of Fe-rich subsurface waters; (ii) from adjacent coastal regions through cross-shelf advection; or (iii) by atmospheric deposition of Fe-rich continental material or from Fe-rich ice melt. Here we are concerned only with physical dynamics that may contribute to mechanisms i and ii.
Eddies and mixing
Vertical mixing has significant implications for biological forcing, including nutrient resupply of the euphotic zone, vertical transport of plankton, photoacclimation, ‘new’ production, etc. (Lewis et al. 1986; Franks et al. 1986; Falkowski et al. 1991; McGillicuddy et al. 1998). To assess the role of vertical mixing in supplying the euphotic zone with Fe, a description of the physical forcings which influence mixing rates, and of mixing rates themselves are necessary. Deep waters, below the euphotic zone, store regenerated nutrients that are available to support photosynthesis if they are transported to the euphotic zone. Eddies are a critical source of energy that can redistribute ocean water both horizontally and vertically. The decay of eddies at large scales can lead to energy at small scales, including greater vertical eddy diffusion coefficients (Kz) within the upper ocean pycnocline. Vertical displacement of the pycnocline, and the associated nutricline induced by open ocean eddy rotation, has been proposed as a significant mechanism that enhances new production in the Sargasso Sea (McGillicuddy et al. 1998). This study found that vertical displacements of the upper ocean pycnocline were correlated to eddy intensity determined from satellite altimetry.
Lateral transport across the shelf may also be an important mechanism for Fe enrichment of the western Scotia Sea. The frontal zone of the Antarctic Circumpolar Current (ACC) is a region with especially strong mesoscale eddy activity. The path of the wind-driven ACC has long been known to be steered by deep seafloor bathymetry (Gordon and Baker 1986). More recently, altimetric investigations have shown that eddy activity in the ACC is also modulated by bottom bathymetry (Sandwell and Zhang 1989, Chelton et al. 1990; Morrow et al. 1992a, 1992b; Gille 1994). Waters less than about 2000 m depth will interact with the current to produce significantly higher eddy energy. Moore et al. (1999a) reported elevated chl-a near the Pacific-Antarctic rise near the US JGOFS survey region which is also evident in Figure 1A.
Enhanced vertical fluxes are likely when internal waves interact with the coastal shelf of South America and the Antarctic Peninsula. Mixing induced by shoaling internal waves can be followed by subsequent intrusions (Thorpe 1982; 1996). Waters with increased nutrient loading would be swept downstream of the Antarctic Peninsula, as has been suggested previously by Sullivan et al. (1993). Similarly, increased turbulent kinetic energy is likely as waters of the Antarctic Circumpolar Current are accelerated through the Drake Passage and through interaction with the Shackleton Fracture Zone which forces the flow of the ACC through several relatively shallow gaps in this mid-ocean rise. Increased baroclinic instability within the PFZ is likely. When such instability has been observed in the Ushant frontal zone near Great Britain, cyclonic eddies form contributing to phytoplankton development (Pingree and Griffiths 1978). Fine-scale temperature variations are a feature of the PF region on kilometer scales (Georgi 1978), and, as discussed by Marmorino et al. (1985) can be a signature of intrusions associated with lateral advection of water. Larger-scale features, such as cyclonic or anti-cyclonic rotation that may be induced along the PF or the shelf breaks bordering the study region may contribute to lateral movement of water off the shelf and across the PF and vertical transport of deep water toward the surface.
When a mean flow encounters bottom bathymetry features, the flow can detour around the bathymetry feature rather than over it. Strong interaction can produce closed circulation and the fluid over the bathymetry feature can even become bathymetrically trapped in what is called a Taylor column (see discussions in Chapman and Haidvogel 1992 and Schär and Davies 1988). Hogg (1980) studied how bathymetric features influence ocean currents. Divergence of flow from the mean path was explained by conservation of potential vorticity (PPV). Therefore mesoscale relief can produce closed streamlines over the bottom features. The water column initially trapped over the feature may either be advected away or remain trapped over the feature (Hogg 1980) depending on the incident flow speed. Chapman and Haidvogel (1992) studied the influence of size, height, shape and roughness of an isolated seamount and the flow stratification on formation of Taylor Caps. Their results indicate that Taylor Caps can generate uplifted or perturbed isopycnals, bottom currents, internal lee waves and vertical water oscillation on the downstream side of the seamount. Alverson (1995) studied the potential for bathymetric features to precondition the open ocean deep convection. He suggested that the bottom bathymetry can cause enhanced oceanic convection which may dictate the location and scale of deep convection. For example, seamounts can cause convective chimneys in weakly stratified deep water and deep eddies can be generated which enhance vertical transport. Thus, when surface flow is influenced by bottom bathymetry, there can be important implications for oceanic circulation, mixing, water property and nutrient distributions, and biological activity.
The ACC and associated fronts are steered by bathymetry at the global scale because the dynamics tend to conserve potential planetary vorticity (PPV) (Moore et al. 1999a, 1999b, 1997). PPV is equivalent to the ratio f/H where f is the Coriolis force and H is bottom depth. Conservation of PPV is consistent with the theoretical developments and oceanic observations of Hogg (1980). The Fine Resolution Antarctic Model (The FRAM Group 1991) resolved variability of the ACC transport, which was shown to follow contours of f/H for most regions (Hughes et al. 1999). Satellite sea surface temperature analysis has shown that the mean path and location of the PF conserves PPV (Moore et. al. 1997, 1999a, 1999b).
These processes are also strong at the mesoscale. The ACC region and Drake Passage may exhibit mesoscale meanders and eddies associated with rises and depressions as the ACC flows from deep water in the south Pacific over the shallower and complex bathymetry of Drake Passage (Moore et al. 1999a, 1999b). For the eastern Drake Passage and the western Scotia Sea regions, which have large bathymetric features, the ACC is forced to cross the f/H contour, which inputs relative vorticity to the water column, resulting in enhanced energy dissipation by eddy action. Moore (1999a) also reported that the PF is intensified here. In fact, this region has the highest temperature change across the front, the widest front width and the largest mesoscale variability among all regions along Antarctic Polar Front. Elevated ocean surface slope variability and intensified mesoscale variability are expected to be found downstream of bathymetric features (Chapman and Haidvogel, 1992; Hogg 1980). In satellite images, many mesoscale meanders and warm and cold core eddies are observed above or downstream of the significant bathymetry features. The weak density stratification and appropriate flow speed in this region may make the bathymetry effect more obvious.
Mesoscale meanders and eddies could strongly influence the local biological environment. Flierl and Davis (1993) suggest that mesoscale meanders can induce along-isopycnal upwelling and downwelling that would bring nutrients into the mixed layer and enhance phytoplankton growth (Hitchcock et al. 1993). Mesoscale eddies can also cause vertical motion (Chapman and Haidvogel 1992, Alverson 1995), which can significantly perturb the plankton population. Many observations of this phenomenon in different regions of the world have been made. A deep cold-core eddy has been documented on Chatham Rise (42.2S and 178.3E), where hydrographic data reveals upwelling and intrusion of water from the south (Chiswell and Sutton 1998). Many cyclonic gyres are observed off the Antarctic Peninsula. Data indicates that local upwelling occurs in these cyclones (Stein 1992; Letelier et al. 1997). Biggs and Müller-Karger (1994) studied the cyclonic-anticyclonic eddy pairs in the western Gulf of Mexico. They reported that the nutrient concentrations and vertically integrated chlorophyll stocks are always higher within cyclones due to upwelling and lower in anticyclones due to downwelling. They also observed that the offshore flow set up along the confluence of the eddies increased chl-a concentration of the offshore waters by dispersion of shelf-borne phytoplankton into the interior of the gulf. The Pacific-Antarctic Ridge is associated with elevated chl-a that Moore et al. (1999a, 1999b) proposed is due to mesoscale physical-biological interactions. When the ACC encounters the ridge, a potentially greater flux of Fe from deeper water upwells, supporting the local phytoplankton crop.
Figure 3A illustrates that the sea surface slope variability is elevated as the ACC encounters the complex bathymetry of Drake Passage. The dramatic demise of the BWZ as the ACC emerges from the Drake Passage is evidence that the physical dynamics have an important influence on the biological dynamics in this region.
The objectives of this research are to define possible physical mechanisms related to fluid flow over bathymetry which might contribute to the disintegration of the BWZ and the elevated chlorophyll of the western Scotia Sea south of the PF.
The hypotheses are:
1. The surface flow of the ACC through the Drake Passage is influenced by bathymetric
rises and depressions.
2. High chlorophyll will be observed downstream of bathymetric features that result in cyclonic, upwelling favorable circulation, a divergence in the flow field or cross-shelf transport from the Antarctic Peninsula or South American shelf regions.
3. Eddies associated with bathymetry contribute to cross-shelf and cross-PF transport which can lead to advection of iron-rich waters from the Antarctic Peninsula shelf and the South American shelf into the western Scotia Sea.
Monthly satellite sea surface temperature data from the AVHRR Pathfinder 4.0 SST global area coverage (GAC) for the period 1987 –1999 were obtained from the NASA Physical Oceanography Distributed Active Archive Center (PODAAC) at NASA’s Jet Propulsion Laboratory. Monthly and 8 day SeaWiFS surface chl-a GAC data (1997 –1999) were obtained from the NASA SeaWiFS DAAC at NASA Goddard Space Flight Center. Both of these data sets are binned to 9 km pixel size. The satellite images were imported to NASA’s SEADAS Software (Fu et al. 1998), which was used to create the maps and to merge with land, latitude, longitude and bathymetry masks.
Data for Lagrangian drifters for the Drake Passage region were provided by P. Niiler of Scripps Institution of Oceanography. The drifters were standard WOCE-style surface drifters drogued at 15 m (Niiler 1995). The data file for each drifter includes latitude, longitude, time, temperature and velocity vectors. Drifter tracks were imported as a mask into SEADAS for display with the bathymetry, SST and chl-a maps.
The bathymetry data available as part of SeaDAS ancillary data files was loaded and projected in SEADAS. The data has 5 minute resolution.
Figure 3A illustrates the large-scale bathymetry of the southeast Pacific Ocean Drake Passage and Scotia Sea. The region is characterized by relatively deep water (4000-6000 m) which rises toward the Drake Passage. Within the Drake Passage, the typical depth is less than 4000 m and there are numerous ridges and rises that are less than 3000 m. The mean surface chl-a derived from SeaWiFS for January (1998 and 1999 composite) exhibits large gradients in this region (Figure 3B). Coastal regions of South America and the Antarctic Peninsula have high chl-a concentrations ranging from 0.5-5.0 mg chla m-3. A prominent low chl-a feature with values less than 0.2 mg chl-a m-3 is evident in the southwestern portion of Figure 3B but abruptly ends at the boundary of the Drake Passage and Scotia Sea. As noted in the introduction, the transition from the blue water zone (BWZ) in the Drake Passage to the high chl-a of the western Scotia Sea is among the strongest gradients in surface chl-a in pelagic waters of the ACC south of the PF. Figure 3C is a map of the ocean surface slope variability derived from the GEOSAT and TOPEX satellite altimetry time series (Yale et al. 1995). The ocean surface slope variability is directly proportional to the eddy kinetic energy multiplied by the Coriolis parameter (f); f varies by about 10% over this region. Regions with high slope variability have high transient eddy energy. The band of surface slope variability is associated with higher chl-a waters along the PF which may be caused by enhanced vertical flux of iron and silicate in the PF zone. The centerline of the BWZ in Figure 3B runs from approximately 66°S and 80°W to 60°S and 60°W. Figure 3C indicates that the core of
the BWZ is a region with relatively low surface slope variability. While the chl-a is generally higher in the band of highest surface slope variability. The eastern boundary of the BWZ at about 58°S and 48°W is not a region with particularly high surface slope variability but is clearly south of the band of high slope variability that follows the general northward flow of the ACC as the current emerges from the Drake Passage along the continental shelf of South America. Since the altimetry data are referenced to long-term average conditions, stationary eddy features, if present, will not result in large slope variability, although they could have significant influence on local dynamics.
Figure 4 is a daily SeaWiFS chl-a GAC image of the region for January 15, 1998 which reveals the strong gradients associated with the boundaries of the BWZ. The northern boundary is the PF, the southern boundary is the Antarctic Peninsula shelf, and the eastern boundary appears to be related to the region of dramatic bathymetric variability. There is significant evidence of eddy features on the boundaries of these transitions above the large topographic features as well as a long-lived high chl-a feature at 60°S and 59°W. This long-lived feature that results in a local high in the 2-year mean chl-a (Figure 5B) is clearly evident in the single day image (Figure 4). The jet like offshore transport of high chl-a shelf water also appears related to the bathymetry. In Figure 4 the inset is the relative concentration of surface chl-a (re-scaled for maximum contrast) for the region in the vicinity of the prominent rise associated with the Shackleton Fracture Zone.
Figure 5A is a map of the Drake Passage region bathymetry which exhibits a series of shallow and deep features between the continental shelves of South America and the Antarctic Peninsula. The 2-year composite SeaWiFS chl-a image (Figure 5B) reveals the BWZ and the relatively high chl-a of the southwestern Scotia Sea. As noted above, the region just north east of the rise at 60°S and 59°W has 2-year mean chl-a values substantially higher than those in the BWZ immediately to the west of this prominent rise.
Three surface drifter tracks in Figure 5B to illustrate that the surface flow is influenced by bathymetry features. The northern and southern drifter tracks generally follow the shelf break of South America and the Antarctic Peninsula, respectively. However, each is clearly influenced by small-scale bathymetric features. The northern buoy completes a full rotation downstream of the Shackleton Fracture zone (57°S 60°W). The drifter in the central Drake Passage begins west of the southern ridge of the Shackleton Fracture zone and is trapped for several revolutions upstream of the rise at 60°S and 59°W before moving east. As discussed above, there is a high in the 2-year chl‑‑a east of the rise, which trapped the drift buoy. This elevated chl-a is also evident in the 2-year January composite of Figure 5B. Its apparent isolation from the high chl-a of the Antarctic Peninsula region and may be the result of local dynamics. The BWZ, which is a massive feature of the South Pacific sector of the Southern Ocean south of the PF, appears to end abruptly in the Drake Passage as the current passes over the bathymetric features evident in Figure 5A.
Figure 6 illustrates the same drifter tracks as in Figure 5B, but overlaid on the monthly SST image that corresponds to the time period the drifter was in the region. These tracks together with SST are particularly interesting with respect to evaluating the possible mechanisms which may lead to the physical dynamics required to increase iron flux to the euphotic zone in the western Scotia Sea. Figure 6A clearly shows that the waters of the Bransfield Strait, which have been shown by Martin et al. (1990) to be enriched with Fe, can be transported into the western Scotia Sea. The drifter follows the shelf break of the peninsula until about 60°S and 51°W where it makes a cyclonic turn downstream of a prominent rise. Following the rise there is deeper water, and the drifter then exhibits an anti-cyclonic turn, and completes several full closed cyclonic rotations downstream of a second prominent rise located at 58°S and 48°W. There appears to be colder water moving north when the drifter moves north near 60°S and 51°W, and warmer water moving south when it undergoes the cyclonic closed circulations near 59°S and 47°W. The drifter on the northern boundary of the region (Figure 6B) also appears to be influenced by the bathymetry, completing a closed rotation at 57°S and 60°W. In addition, the warm water seems to move south from the PF to the large depression at 56°S and 51°W. The drifter that was released in the central domain of the BWZ at about 60°S and 60°W initially becomes trapped on a rise for 46 days completing 4 anti-cyclonic rotations before moving across the Drake Passage. Upstream of the long rise that extends from 59°S and 60°W to 61°S and 57°W, the drifter is steered poleward, then made an anti-cyclonic meander following the corner of the large depression at 56°S and 51°W. This southerly meander appears to pull warmer water from the north toward the south.
While it is impossible to interpret the complex flow of this region from the behavior of so few surface drifters, the patterns of motion are consistent with the mechanisms which we believe to have an important impact on vertical and lateral transport of iron into the western Scotia Sea. Specifically, cross shelf transport is evident in Figure 6A, and all the drifters exhibit cyclonic and anti-cyclonic motions that are consistent with the predictions of the flow direction required to conserve PPV.
Figure 7A is a monthly time series for 1998 of satellite-derived SST at 58°S and 54°W compared to SST at the same latitude but to the east (52°W and 51°W). A prominent elevation of SST is evident at 54°W from July through November 1998 compared to the locations to the east. The October 1998 SST image (Figure 7B) indicates that this divergence in surface temperature is associated with a warm water eddy. This eddy has been observed in most monthly GAC SST imagery each year for the September-November period (data from 1987-1999 was examined but not shown here). The feature appears to be a quasi-stationary warm eddy associated with a deep bathymetric feature (see bathymetry in Figure 5A). The warm feature appears isolated from the warmer water of the PF for the image in 7B. The dynamics of its formation and maintenance are of interest. It may originate as a meander in the mean flow of the PF, and become bathymetrically trapped. Figure 7C shows that the warm feature corresponds to a low in chl-a, compared to the relatively high chl-a that propagates east of 59°S and 58°W associated with the rise discussed above (Figure 4). The elevated SST feature at 58°S and 54°W in Figure 7B and the elevated chl-a at 59°S and 58°W are features that recur at the same location in different months and different years.
The imagery is consistent with quasi-stationary circulation trapped by bathymetry, with anti-cyclonic rotation downstream of the depression at 58°S and 54°W and cyclonic rotation downstream of the rise at 60°S and 58°W. If the flow downstream of the rise at 60°S and 59°W is cyclonic, there may be localized upwelling which could transport limiting Fe to the surface (Letelier et al. 1997). The warm water from the South American shelves appears transported into the central region of the eastern Drake Passage along depressions, while the cold high chl-a water from the Antarctica Peninsula shelves is also transported into the central region of the eastern Drake Passage along the rise. The temperature gradient across the front can be very high (> 6°C).
Figure 8A and 8B provide evidence of warm core and cold core eddies in the PF region that may be associated with bathymetric interactions. The lateral and vertical flow in the vicinity of these features may transport water from the coastal shelf of South America toward the center of the Drake Passage, cold water from the south into the core of the ACC jet, or deep water to the surface. These dynamics could re-distribute Fe in the region contributing to the demise of the BWZ and the elevated chl-a in the southwestern Scotia Sea. The lateral transport from south to north possible by the eddies would be consistent with the large cross-PF transport of ALACE Floats reported by Davis et al. (1996).
Oceanographers have been aware of the importance of eddy motions in the ocean since the MODE and POLYMODE programs in the 1970s. Bryden (1983) provided an overview of early research on Southern Ocean eddies. He noted three ways of characterizing eddies: (1) meridional displacements of zonal jets, (2) cyclonic or anticyclonic rings that separate from the jets, and (3) wave-like meanders of the jets which propagate along the jet but do not separate.
Foster and Middleton (1984) carried out a study of water mass mixing in the study region with a detailed hydrographic survey with relatively high spatial resolution from 50°W to 35°W and from 60°S to 52°S. Their westernmost sections were rather short and restricted to the southern half of the latitude domain. Their results suggested strong mesoscale eddy features with typical length scales of less than 75 km. Analysis of potential temperature – salinity diagrams (T-S) indicated strong mesoscale mixing of Drake Passage, Weddell Sea and Bransfield Strait water although the former two water types were found to be dominant. The southern-most stations along 50°W (between 60-61°S) exhibited typical Weddell Sea characteristics; water farther north along the same line appeared to be a complex mixture of Drake Passage and Weddell Sea water. Farther east, there was evidence of northward transport of relatively coherent Weddell Sea water that was found north of water with typical Drake Passage characteristics. The eastern domain appeared to have significant eddy-like features and water masses became less defined due to apparent mixing. Earlier studies described evidence of significant mixing
of water masses in the region, including ACC, Bellingshausen Sea, Bransfield Strait and Weddell Sea waters (Deacon, 1937; Patterson and Sievers, 1980). Deep mixing has also been reported in the vicinity of the Scotia Ridge that rises to about 600 m along 61°S in the vicinity of 45-50°W (Deacon and Foster, 1977; Michel, 1984) which often corresponds to a frontal boundary between Weddell Sea and Drake Passage water (Maslennikov and Solyankin, 1979; Pattterson and Sievers, 1980). The hydrographic literature indicates that the region of study is characterized by strong mixing of different water types with significant evidence of mesoscale eddies.
Eddies and rings that form from strong currents have been most closely studied near the Gulf Stream. Southern Ocean eddies differ from Gulf Stream eddies in a number of important ways. First, due to the higher latitude and smaller Rossby radius of deformation, Southern Ocean eddies tend to be smaller, with radii typically estimated between 30 and 50 km (Bryden 1983), in contrast to values of 45 to 75 km for Gulf Stream warm core rings (Brown et al. 1986; Cornillon et al. 1989). Second, because of the relatively low stratification in the Southern Ocean, eddies tend to be vertically coherent from the surface to the bottom (Bryden 1983) and may cause deep vertical convection (Alverson 1995). Third, flow speed in waters of the Southern Ocean is slower than the Gulf Stream (maximum velocity of about 1m/s), therefore eddies may be trapped on the topography feature instead of advected away downstream (Hogg 1980).
Like Gulf Stream rings, some of the Southern Ocean rings in this study appear to form when large meanders break away from the Polar Front. The positions of Southern Ocean eddies in the Drake Passage region appear to be closely linked to rises and depressions in the seafloor. In contrast, the Gulf Stream is more strongly stratified than the Southern Ocean and flows over relatively smooth and flat bathymetry. Studies of Gulf Stream rings and eddies do not specifically report that bathymetry has any influence on ring formation locations (Brown et al. 1986; Cornillon et al. 1989; Lee and Cornillon 1996) although there is ample evidence that dynamical processes associated with cyclonic flow result in substantial upwelling of nutrients (Hitchcock et al. 1993; McGillicuddy et al. 1998).
Once Gulf Stream rings form, they tend to propagate westward. Warm core rings are estimated to have a propagation speed of 6 to 8 cm/s southwestward over their lifetimes (Brown et al. 1986), but over short time periods may propagate northwestward (Cornillon et al. 1989). Warm core Gulf Stream rings have an average lifetime of 130 days and may last up to 400 days. They disappear either by coalescing with the Gulf Stream or by decaying with an e-folding scale of 213 days (Brown et al. 1986). In contrast, some of the Southern Ocean rings observed in this study appear to be trapped due to bathymetry, for extended periods of time.
The bathymetric influence on the Antarctic Circumpolar current (ACC) has been studied previously (Moore et al. 1997, 1999a, 1999b). The mean flow of the ACC is from west to east with local maximum velocities located along the PF. The Polar Front has a surface velocity estimated between 23cm/s and 40cm/s from floats and hydrographic data (Whitworth and Nowlin, 1987). Davis et al. (1996) reported mean current flows determined from deep drifters (ALACE) in the northern Drake Passage of approximately 12-14 cm s-1. They also reported major cross-ACC transport, with one float moving from the north to the south and another moving from the south to the north with transports of up to 200 km across the mean flow of the ACC. Their results implied very dramatic mesoscale transport for this region, which may be related to the interaction of the ACC with the bathymetry and the mesoscale eddies we describe. Moore (1999a) also reported that the PF is intensified here. In fact, this region has the highest temperature change across the front, the widest front width and the largest mesoscale variability among all regions along Antarctic Polar Front. Figure 3C also shows that this region has greater sea surface slope variability associated with eddies.
When the ACC passes the narrow Drake Passage it interacts with the complex bottom bathymetry in the Drake Passage and Scotia Sea. Due to the tendency to conserve planetary potential vorticity (PPV = f/H), when the flow passes a depression on the sea floor, the increased (H) causes the flow to move poleward so that f will increase (f=2WSinF; W is the angular rotation of the Earth on its axis and is a constant, F is the latitude) followed by a downstream anti-cyclonic meander. Conversely, when the flow encounters a rise, H decreases and the flow tends to move toward the Equator to conserve PPV with a subsequent downstream cyclonic meander.
Figure 9 is an idealized diagram of the possible flow downstream of bathymetric features. Cyclonic rotation may be associated with a rise and anti-cyclonic rotation may be associated with a depression (Figures 9A and 9B). The current may be steered south downstream of rises, and north downstream of depressions. An anti-cyclonic meander of the main flow along the Polar front caused by a depression may transport warm water from north of the front and trap it as a warm eddy associated with the depression. A cyclonic meander near a rise can create a cold core eddy by moving cold water from south of the PF into warm water to the north. Examples of surface SST that are consistent with these mechanisms are shown in Figure 8. This dipole-type eddy feature would be consistent with the hypothesized mechanism diagrammed in Figure 9B. The region of the eastern Drake Passage bounded by 56°S and 56°W on the northwest corner and 60°S and 44°W on the southeast corner has a series of rises on the southern side, and a series of depressions on the northern side. As suggested by the drifter tracks in Figure 5B, there may be a tendency for surface flow on the southern side of this region to move towards the north in cyclonic meanders, and on the northern side of this region to move towards the south in anti-cyclonic meanders. This mechanism is represented conceptually by Figure 9C. The interaction between these cyclonic and anti-cyclonic meanders may cause local divergences and convergences in this zone, thus producing local vertical motion and phytoplankton perturbation. The offshore movement may also transport high chl-a shelf water to the center of the flow. In addition, the divergence created when shelf-waters move offshore may increase the nutrient upwelling over the continental margin, leading to the high chl-a concentration in the shelf-water (Biggs and Müller-Karger 1994). If this type of flow dynamics does exist in the eastern Drake Passage, one predicts upwelling in the divergence zone, which may elevate surface Fe and enhance surface chl-a in the western Scotia Sea. However, the formation of eddies also depends on many other factors (Hogg 1980). Even in the absence of bathymetric influence, eddies would form on the PF due to baroclinic instability, which has been estimated to be significant in the Drake Passage region (Wright 1981).
Interestingly, the drifter released just west of the Shackleton Fracture Zone rise at approximately 60°S and 60°W (Figure 6C) becomes trapped in an anti-cyclonic eddy for 46 days upstream of the ridge and completes 4 full rotations. The average radius of the eddy over the 4 cycles was quite small (28 km), consistent with earlier studies of typical eddy scales (Bryden 1983). Apparently this drifter moved north, as expected upon encountering the ridge system, but because the ridge is long and oriented toward the northwest (into the mean flow of the current), the drifter could not escape downstream. It was repeatedly pushed back onto the ridge by the prevailing easterly flow, then was forced northwest into the mean flow as it approached the rise each subsequent cycle. Eventually it reached the northwest corner of this long ridge, then made a sharp cyclonic turn to the east. Such movement against the mean flow has been noted for ALACE floats by Davis et al. (1996). Clearly, there is no simple model that can account for the complexity of the fluid dynamics of this (or any other) real system. Idealized models do not incorporate the complexity of the real system. Eddies formed by bathymetric mechanisms can be transient and transported downstream from the location where they are generated. The probability of an eddy becoming trapped on a bathymetric feature is greater when the mean flow is low (Hogg 1980). Typically, the rotation predicted for an idealized bathymetric feature is downstream of the feature (Hogg 1980; Chapman and Haidvogle 1992). In the case of the drifter in Figure 6C, the initial tendency of the float to move north upon encountering the rise is predicted. However, in this case, rather than observing a cyclonic rotation downstream, an anti-cyclonic eddy formed upstream and persisted for an extended period of time.
The existence of mesoscale meandering and eddy features can cause local upwelling and downwelling (Letelier et al. 1997). Cyclonic rotation is associated with divergence and upwelling (Stein 1992; Chiswell and Sutton 1998). Seamounts can produce perturbed isopycnals and vertical movement (Chapman and Haidvogel 1992; Alverson 1995; Moore et al. 1999a). Mesoscale meanders may cause local upwelling (Flierl and Davis 1993; Hitchcock et al. 1993).
Eddies and meanders of the ACC in the study region will influence the distribution of physical and chemical properties (McGillicuddy et al. 1998). For example, upwelling can increase vertical mixing bringing deep iron to surface waters, thus enhancing phytoplankton biomass within cyclonic eddies (Chiswell and Sutton 1998; Letelier et al. 1997), above seamounts (Moore et al. 1999a, 1999b) and near the mesoscale meanders (Flierl and Davis 1993; Hitchcock et al. 1993). Moreover, lateral cross-shelf transport may bring iron from the Antarctic Peninsula shelf or the South America shelf into the western Scotia Sea, thereby contributing to the termination of the BWZ. North of the PF, vertical or lateral transport of silicate will be important for diatoms that are prevalent in this region. In addition, the divergence created when shelf-waters move offshore may increase nutrient upwelling over the continental margin, and lead to the high chl-a concentration of the shelf-water.
In summary, the Drake Passage and western Scotia Sea have many bathymetric features compared to the smooth and deep ocean bottom to the west where the BWZ appears to maintain its integrity each year. The data presented here are consistent with mechanisms of bathymetry-induced mesoscale dynamics that lead to re-distribution of nutrients required for phytoplankton growth. From satellite images and drifter data, many mesoscale meanders and eddies are observed above or downstream of these bathymetric features. Previously published models support that these mesoscale meanders and eddies may generate vertical mixing and lateral movement, thus influencing the distribution of physical and chemical properties. Furthermore, elevated chl-a is observed downstream of these bathymetric features, which may be due to the nutrient input by local upwelling and offshore transport associated with topographically trapped eddies and meanders. In addition, krill may be supported by the biomass in this high chlorophyll water, or may be transported and trapped by these mesoscale meanders and eddies. Therefore, the BWZ appears to end on this large topography region and high chlorophyll water and high krill population appears to start here.
Satellite-derived chl-a maps of the Southern Ocean have revealed a massive blue water zone (BWZ) south of the PF in the south Pacific region. This BWZ never has chl-a > 0.2 mg m-3 and has peak chl-a in October, rather than December as in other regions with comparable proximity south of the PF. The BWZ ends abruptly as the ACC flows through the Drake Passage into the Scotia Sea. Recurring features in surface chl-a and SST indicate that the BWZ ends as the ACC flows over prominent bathymetric depressions and rises in the Drake Passage. Previously published models support the concept that the flow of the ACC over bathymetric features could produce vertical and lateral motion associated with stationary eddy features and mesoscale meanders. The drifter and satellite data presented here provide evidence that supports each of the hypotheses regarding mixing (vertical and lateral) that were evaluated. However, the amount of drifter and hydrographic data is still too sparse to prove that the various mechanisms discussed here are in fact so influential. Still, these results are fully consistent with the overall hypothesis that Fe that accumulates in deep water through re-mineralization, or that is entrained as deep water passes over the shelf of South America or the Antarctic Peninsula, can be transported to the euphotic zone of the southwestern Scotia Sea. The fact that the terminus of the BWZ is only tens of kilometers from both the lowest and highest chl-a biomass regions of the Southern Ocean south of the PF implies that there must be physical mechanisms for near-surface iron enrichment. Further understanding will require detailed hydrographic surveys with high spatial resolution (or using continuous profilers like SeaSoar) combined with dynamical studies of vertical and
lateral transport, characterization of the Fe concentrations in source and surface waters of interest, and documentation of phytoplankton physiology and growth response to the physical-chemical forcing.
Biggs, D.C. and F.E. Müller-Karger (1994) Ship and satellite observations of chlorophyll stocks in interacting cyclone-anticyclone eddy pairs in the western Gulf of Mexico. Journal of Geophysical Research. 99(C4): 7,371-7,384
Brown, O.B., P.C. Cornillon, S.R. Emmerson and H.M. Carle (1986) Gulf Stream warm rings: a statistical study of their behavior. Deep-Sea Res.Part A-Oceanographic. 33: 1,459-1,473
Bryden, H. (1983) The Southern Ocean. In: Eddies in Marine Science, Robinson, A. R., ed. 265-277
Buma, A.G.J., H.J.W. de Baar, R.F. Noltig and A.J. van Bennekom (1991) Metal enrichment experiments in the Weddell-Scotia Seas: Effects of iron and manganese on various plankton communities. Limnology and Oceanography. 36(8): 1865-1878
Chapman, D.C. and D.B. Haidvogel (1992) Formation of Taylor caps over a tall isolated seamount in a stratified ocean. Geophys.Astrophys.Fluid Dynamics. 64: 31-65
Chelton, D.B., M.G. Schlax, D.L. Witter and J.G. Richman (1990) Geosat altimeter observations of the surface circulation of the Southern Ocean. Journal of Geophysical Research. 95: 17,877-17,903
Chisholm, S.W. and F.M.M. Morel (1991) What controls phytoplankton production in nutrient-rich areas of the open sea? Limnology and Oceanography. 36(8): U-1,507-U-1,511
Chiswell, S.M. and P.J.H. Sutton (1998) A Deep Eddy in the Antarctic Intermediate Water North of the Chatham Rise. Journal of Physical Oceanography. 28(3): 535-540
Cornillon, P., R. Weyer and G.R. Flierl (1989) Translational velocity of warm core rings relative to the slope water. Journal of Physical Oceanography. 19: 1,317-1,332
Davis, R.E., P.D. Killworth and J.R. Blundell (1996) Comparison of autonomous lagrangian circulation explorer and fine resolution Antarctic model results in the South Atlantic. Journal of Geophysical Research. 101(C1): 855-884
de Baar, H.J.W., A.G.J. Buma, R.F. Nolting, G.C. Cadee, G. Jacques and P.J. Treguer (1990) On iron limitation of the Southern Ocean: experimental observations in the Weddell and Scotia Seas. Mar Ecol Prog Ser. 65: 105-122
Deacon, G.E.R. (1937) The hydrology of the Southern Ocean. Discovery Reports. 15: 1-24
Deacon, G.E.R. and T.D. Foster (1977) The boundary region between the Weddell Sea and Drake Passage currents. Deep-Sea Research. 24: 505-510
El-Sayed, S.Z. (1987) Biological production of Antarctic waters: Present paradoxes and emerging paradigms. In: Antarctic aquatic biology, BIOMASS Scientific Series, Volume 7, SCAR, ed. 1-21
Falkowski, P.G., D. Ziemann, Z. Kolber and P.K. Bienfang (1991) Role of eddy pumping in enhancing primary production in the ocean. Nature. 352: 55-58
Flierl, G.R. and C.S. Davis (1993) Biological effects of Gulf Stream meandering. Journal of Marine Research. 51: 529-560
Foster, T.D. and J.H. Middleton (1984) The oceanographic structure of the eastern Scotia Sea-I. Physical oceanography. Deep-Sea Research I. 31: 529-550
Franks, P.J.S., J.S. Wroblewski and G.R. Flierl (1986) Prediction of phytoplankton growth in response to the frictional decay of a warm-core ring. Journal of Geophysical Research. 91(C6): 7,603-7,610
Froneman, P.W., R. Perissinotto, C.D. McQuaid and R.K. Laubscher (1995) Summer distribution of net phytoplankton in the Atlantic sector of the Southern Ocean. Polar Biology. 15: 77-84
Fu, G., K. Settle and C.R. McClain (1998) SeaDAS: The SeaWiFS Data Analysis System. Proceedings of the Fourth Pacific Ocean Remote Sensing Conference, Qingdao, China.73-77. July 28, 1998
Georgi, D.T. (1978) Fine structure in the Antarctic Polar Front Zone: Its characteristics and possible relationship to internal waves. Journal of Geophysical Research. 83: 4,579-4,588
Gille, S.T. (1994) Mean surface height of the Antarctic Circumpolar Current from Geosat data: Method and application. Journal of Geophysical Research. 99: 18,255-18,273
Gordon, A.L. and T.N. Baker (1986) Southern Ocean Atlas. International Decade of Ocean Exploration.
Hays, P.K., T.M. Whitaker and G.E. Fogg (1984) The distribution and nutrient status of phytoplankton biomass in the Southern Ocean between 20° and 70°W. Polar Biology. 3: 153-165
Helbling, E.W., V.E. Villafañe and O. Holm-Hansen (1991) Effect of iron on productivity and size distribution of Antarctic phytoplankton. Limnology and Oceanography. 36: 1879-1885
Hitchcock, G.L., A.J. Mariano and T. Rossby (1993) Mesoscale pigment fields in the Gulf Stream: Observations in a meander crest and trough. Journal of Geophysical Research. 98: 8,425-8,445
Hofmann, E.E. and J.M. Klinck (1998) Thermohaline variability of the waters overlying the west Antarctic Peninsula continental shelf. In: Ocean, Ice and Atmosphere: Interactions at the Antarctic Continental Margin, Jacobs, S. S. and R. Weiss, eds. Edition 75, 67-81
Hogg, N.C. (1980) Effects of bottom topography on ocean currents. In: Global atmospheric research programme, Joint Organizing Committee167-205
Hughes, C.W., M.P. Meredith and K.J. Heywood (1999) Wind-driven transport fluctuations through Drake Passage: a southern mode. Journal of Physical Oceanography. 29(8): 1,971-1,992
Lee, T. and P. Cornillon (1996) Propagation and growth of Gulf Stream Meanders between 75° and 45°W. Journal of Physical Oceanography. 26: 225-241
Letelier, R.M., M.R. Abbott and D.M. Karl (1997) Chlorophyll natural fluorescence response to upwelling events in the Southern Ocean. Geophysical Research Letters. 24(4): 409-412
Lewis, M.R., W.G. Harrison, N.S. Oakey, D. Herbert and T. Platt (1986) Vertical nitrate fluxes in the oligotrophic ocean. Science. 234: 870-873
Loscher, B.M., H.J.W. de Baar, J.T.M. de Jong, C. Veth and F. Dehairs (1997) The distribution of Fe in the Antarctic Circumpolar Current. Deep-Sea Research. 44(1-2): 143-187
Marmorino, G.O., L.J. Rosenblum, J.P. Dugan and C.Y. Shen (1985) Temperture fine-structure patches near an upper ocean density front. Journal of Geophysical Research. 90: 11,799-11,810
Martin, J.H., R.M. Gordon and S.E. Fitzwater (1990) Iron in Antarctic waters. Nature. 345: 156-158
Maslennikov, V.V. and Y.V. Solyankin (1979) Interannual displacements of the zone of interaction of Weddell Sea waters with the Antarctic Circumpolar Current. Antarktika. 18: 118-122
McGillicuddy, Jr., D.J., D.A. Siegel, H.W. Jannasch, R. Johnson, T.D. Dickey, J. McNeil, A.F. Michaels and A.H. Knap (1998) Influence of mesoscale eddies on new production in the Sargasso Sea. Nature. 394: 263-266
Michel, R.L. (1984) Oceanographic structure of the eastern Scotia Sea: II. Chemical oceanography. Deep-Sea Res.Part A-Oceanographic. 31(10): 1,571-1,168
Moore, J.K., M.R. Abbott and J.G. Richman (1997a) Variability in the location of the Antarctic Polar Front (90°-20°W) from satellite sea surface temperature data. Journal of Geophysical Research. 102(C13): 27,825-27,833
Moore, J.K., M.R. Abbott and J.G. Richman (1999b) Location and dynamics of the Antarctic Polar Front from satellite sea surface temperature data. Journal of Geophysical Research. 104(C2): 3,059-3,073
Moore, J.K., M.R. Abbott, J.G. Richman, W.O. Smith, J.R. Cowles, K.H. Coale, W.D. Gardner and R.T. Barber (1999) SeaWiFS satellite ocean color data from the Southern Ocean. Geophysical Research Letters. 26(10): 1,465-1,468
Morrow, R., J. Church, R. Coleman, D. Chelton and N. White (1992) Eddy momentum flux and its contribution to the Southern Ocean momentum balance. Nature. 357: 482-484
Morrow, R., R. Coleman, J. Church and D. Chelton (1992) Surface eddy momentum flux and velocity variances in the Southern Ocean from Geosat altimetry. Journal of Physical Oceanography. 24: 2,050-2,071
Mosby, H.A. (1934) The waters of the Atlantic Antarctic Ocean 1927-1928. Scientific Research of the Norwegian Antarctic Expeditions. 11: 1-131
Niiler, P.P., A.S. Sybrandy, K. Bi, P.M. Poulain and D. Bitterman (1995) Measurements of the water-following capability of holey-sock and TRISTAR drifters. Deep-Sea Research I. 42(11/12): 1,951-1,964
Nowlin, W.D., Jr. and M. Clifford (1982) The kinematic and thermohaline zonation of the Antarctic Circumpolar Current at Drake Passage. Journal of Marine Research. (Supplement) 40: 481-507
Patterson, S.L. and H.A. Sievers (1980) The Weddell-Scotia Confluence. Journal of Physical Oceanography. 10: 1,584-1,610
Pingree, R.D. and W.G. Griffiths (1978) Tidal fronts on the shelf seas around the British Isles. Journal of Geophysical Research. 83: 4,615-4,622
Sandwell, D.T. and B. Zhang (1989) Global Mesoscale Variability from Geosat Exact Repeat Mission: Correlation with Ocean Depth. Journal of Geophysical Research. 94(C12): 17,971-17,984
Schär, C. and H.C. Davies (1988) Quasi-geostrophic stratified flow over isolated finite amplitude topogoraphy. Dynamic Atmosphere Oceans. 11: 287-306
Stein, M. (1992) Variability of local upwelling off the Antarctic Peninsula, 1986-1990. Archiv.fur Fischereiwssenschaft. 41(2): 131-158
Sullivan, C.W., K.R. Arrigo, C.R. McClain, J.C. Comiso and J. Firestone (1993) Distributions of plytoplankton blooms in the southern ocean. Science. 262: 1,832-1,837
The FRAM Group (1991) An eddy-resolving model of the Southern Ocean. EOS Transactions American Geophysical Union. 72(15): 169-184
Thorpe, S.A. (1982) On the layers produced by rapidly oscillating a vertical grid in a uniformly stratified fluid. Journal of Fluid Mechanics. 124: 391-409
Thorpe, S.A. (1996) The cross-slope transport of momentum by internal waves generated by alongslope currents over topography. Journal of Physical Oceanography. 26: 191-204
Timmermans, K.R., M.A. van Leeuwe, J.T.M. de Jong, R.M.L. McKay, R.F. Nolting, H.J. Witte, J. van Ooyen, M.J.W. Swagerman, H. Kloosterhuis and H.J.W. de Baar (1998) Iron stress in the Pacific region of the Southern Ocean: evidence from enrichment bioassays. Marine Ecology Progress Series. 166: 27-41
Whitworth III, T. and W.D. Nowlin, Jr. (1987) Water masses and currents of the Southern Ocean at the Greenwich Meridian. Journal of Geophysical Research. 92(C6): 6,462-6,476
Whitworth III, T., A.H. Orsi, S.-J. Kim, W.D. Nowlin, Jr. and R.A. Locarnini (1998) Water masses and mixing near the Antarctic Slope Front. In: Ocean, Ice and Atmosphere: Interactions at the Antarctic Continental Margin, Jacobs, S. S. and R. F. Weiss, eds. Edition 75, 1-27
Wright, D.G. (1981) Baroclinic instability in Drake Passage. J.Phys.Oceanogr. 11: 231-246
Yale, M.M., D.T. Sandwell and W.H.F. Smith (1995) Comparison of along-track resolution of stacked Geosat, ERS 1, and TOPEX satellite altimeters. Journal of Geophysical Research. 100(B8): 15,117-15,127