Working Group Members: Jack Barth, Mark Brzezinski, Chris Sabine, John Southon, Taro Takahashi
Major Features of the Southern Ocean
The Southern Ocean, defined for the purposes of this study as the region south of, and including, the Subtropical Convergence, covers nearly 20% of the global ocean area. The Antarctic Circumpolar Current has the largest volume flux of any major ocean current (~130 Sverdrups). It is the only continuous circumglobal current, without beginning or end, and it is responsible for mixing of the deep waters of the other major oceans. Because of the absence of land masses to impede its flow, a large component of the flow is barotropic. Most of the ventilation of deep-sea water masses takes place in the Southern Ocean; that is, deep water masses exchange gaseous components, including CO2, with the atmosphere. Furthermore, most deep waters derive their physical, chemical, and biological characteristics in the regions of the Southern Ocean where isopycnals outcrop at the sea surface and where mixing, cooling, and sea ice formation produce new water masses which sink into the ocean interior and renew the intermediate and deep waters of the world's oceans.
Associated with deep-sea ventilation is a series of frontal systems that encircle Antarctica. From north to south the major fronts include the Subtropical Convergence, the Subantarctic Front, and the Antarctic Convergence (Polar Front). These three fronts occur under permanently ice-free conditions, and represent the sites of formation of mode waters in the north and Antarctic Intermediate Water in the south. Further south, under conditions of seasonal ice cover, deep waters are drawn to the surface to the north of the South ACC Front, the southern boundary of the ACC, bringing with them excess metabolic CO2 and inorganic nutrients - the products of centuries of deep-sea respiration in regions to the north. Finally, the densest bottom waters, spreading out over the world's oceans, originate from restricted areas near the Antarctic coast, primarily in the Weddell and Ross Seas.
The extensive regular seasonal advance and retreat of sea ice, oscillating between a maximum coverage of ~20 (10)6 km2 and a minimum of ~4 (10)6 km2, represents the largest seasonal signal of changing environmental boundary conditions in the global ocean. This surface feature, too, can be thought of as a frontal system, one that migrates north and south many hundreds of km annually. Biological productivity of surface waters is strongly influenced by the presence, and melting, of sea ice. Ice-edge productivity supports an abundance of life at higher trophic levels, including mammals and birds as well as zooplankton and fish.
The Southern Ocean is the ocean's largest High-Nutrient Low-Chlorophyll region. Roughly 90% of the phosphate and nitrate is ocean surface waters resides in the Southern Ocean. The potential for altering the partitioning of CO2 between the atmosphere and the deep sea as the result of a perturbation of the efficiency of nutrient utilization in the Southern Ocean has been widely publicized. Some have even constructed schemes to fertilize the Southern Ocean with iron as a geoengineering strategy to reduce the level of anthropogenic CO2 in the atmosphere. The ecological consequences of such an endeavor have yet to be explored.
An organized pattern of atmospheric circulation over the Southern Ocean known as the "Annular Mode" has recently been recognized. Broadly analogous to the North Atlantic Oscillation, the Annular Mode (AM) is characterized by changes in atmospheric pressure gradients between high and temperate latitudes. Wind patterns associated with the AM may exhibit teleconnections to the tropics, and may further explain some of the interannual variability in the Southern Ocean associated with the Antarctic Circumpolar Wave (ACW). The ACW is a wave-number two feature that circles Antarctica with a period of about 8 years, and represents systematic changes in air pressure, air temperature, meridional wind stress, and sea ice extent. Neither the nature of the teleconnections between these features and the tropics, nor the implications for ocean ecology and carbon fluxes, have been studied quantitatively.
Paleoclimate records show strong correlations between environmental conditions in the Southern Ocean and changes in the CO2 content of the atmosphere, as recorded in air bubbles trapped in the polar ice caps. Past atmospheric CO2 concentrations are correlated not only with air temperature over Antarctica, but with sea surface temperature, sea ice extent, and the flux of dust from the atmosphere as well. These correlated features precede the manifestations of climate change in other parts of the world, implicating processes within the Southern Ocean as causal factors regulating natural changes in atmospheric CO2. Because of the unique physical, chemical and biological features of the Southern Ocean described above, perturbation of physical and biogeochemical processes in the region by future climate change may substantially alter the partitioning of CO2 between the ocean and the atmosphere. While the potential for significant feedbacks in the Southern Ocean in response to the rise in anthropogenic CO2 is well recognized, our understanding of the processes in the Southern Ocean regulating carbon fluxes, and of their sensitivity to perturbation, remains insufficient to predict with any confidence the consequences of global warming.
Critical Processes in the Southern Ocean
Superimposed on the large-scale barotropic flow of the ACC is an abundance of mesoscale activity, including eddies, the intensity of which is regulated to some extent by bottom topography. Eddies contribute significantly both to the meridional transport of heat and nutrients as well as to the vertical fluxes of limiting nutrients such as iron and, in some cases, silicon. Meandering of the major frontal features in the Southern Ocean represents another form of mesoscale variability. Vertical displacement of waters influenced by meandering fronts can be on the order of tens of meters per day. Vertical transport of this magnitude affects both the supply of nutrients from below as well as the light regime experienced by phytoplankton.
Although major nutrients such as N and P are rarely depleted from surface waters, nutrient limitation by Fe and, sometimes, by Si as well as by other essential elements, is believed to be common. The implications for ecosystem structure and, hence, for carbon fluxes of altering the supply of Fe and other limiting nutrients is only beginning to be explored.
Changes in stratification of the Southern Ocean, both through natural seasonal and interannual variability of physical conditions as well as due to climate change, are likely to have substantial impacts on carbon fluxes. South of the Polar Front, upper water column stratification is largely determined by cold water of reduced salinity residing over warmer, saltier circumpolar deep water. Warm salty deep waters derive their characteristics by mixing with the deep waters of the Atlantic, Pacific and Indian Oceans. Ventilation of deep waters in the Southern Ocean is directly dependent on the rate of exchange between surface waters and deep waters and this, in turn, is dependent on the strength of the thermohaline stratification of the upper water column. The static stability between the mixed layer and deep water is quite weak, and small perturbations of the salinity budget can upset the stability, thereby leading to a change in the rate of deep convection. Sea ice is maintained in a metastable state through a network of negative feedbacks. Perturbation of thermohaline stratification, and hence of the dependent features of the Southern Ocean such as sea ice distributions and deep-water ventilation, by greenhouse-gas-induced global change will depend, then, in large part on the effect of global change on the fresh water balance of the Southern Ocean.
Paleoclimate records indicate that the mean position of the westerlies shifted meridionally in response to past climate changes, and it is conceivable that future climate change will likewise induce shifts in the position of the winds. Surface wind stress and Ekman divergence are tied to the zero-line of wind stress curl, so it would be expected that these features would shift with any change in the winds. The major fronts, on the other hand, are believed to be locked into their position through interaction with bottom topography. The potential impact on carbon fluxes of a change in the physical relationship between the zero-line of wind stress curl and the Antarctic Polar Front have not been investigated. In addition, regional wind patterns are known to follow the Antarctic Circumpolar Wave as it migrates around Antarctica. Divergence of the westerlies from seasonal sea ice boundaries may well have impacts on ocean ecology and carbon fluxes, although these, too, have yet to be explored.
Anticipated Perturbations of the Modern Carbon Cycle
With increased levels of atmospheric CO2, there will be less need for phytoplankton to pump bicarbonate ions to fill their need for CO2 in support of photosynthesis. This effect is likely to be less significant in the Southern Ocean compared to lower latitudes, however, because the low temperatures of Southern Ocean waters lead to higher concentrations of aqueous CO2, at a given level of pCO2, reducing the potential severity of CO2 limitation during phytoplankton blooms.
A general concern is that the acid titration of the ocean by rising atmospheric CO2 will endanger calcium carbonate-secreting organisms through the reduction of the concentration of dissolved carbonate ion. A widely-held view is that carbonate-secreting plankton are rare in the Southern Ocean, and that one need not consider this perturbation for this region. However, the analysis of sinking particles collected by sediment traps deployed by the U.S. JGOFS Southern Ocean program showed that the export flux of CaCO3 in the Southern Ocean, throughout the region north of the SACCF, is nearly as large as in the central equatorial Pacific Ocean. Ecological changes resulting from reduced carbonate ion concentration in the Southern Ocean cannot be neglected.
General circulation models have predicted an increased flux of fresh water to polar latitudes as a consequence of global warming. As noted above, stratification of the Southern Ocean is weak and precariously balanced. A minor perturbation of the freshwater budget could lead to a radical alteration of the nature and rates of convection and deep water formation, as well as to changes in sea ice extent and seasonality (see below).
The abundance of major nutrients that exists in surface waters of the Southern Ocean reflects the limited capacity of phytoplankton to consume these nutrients due to other factors, such as the availability of iron. Both positive and negative impacts of dust (Fe) supply to the Southern Ocean, as the result of global warming, can be envisioned. Increased temperatures over land may lead to reduced soil moisture and, hence, to increased supply of dust. On the other hand, increased water vapor transport from warmer tropics to the poles will lead to enhanced washout of dust, thereby limiting the effectiveness of poleward atmospheric transport of dust. The net impact of global warming on the supply of dust to the Southern Ocean remains unknown.
Seasonal advance and retreat of sea ice over the Southern Ocean represents one of the largest sources of environmental variability on earth. Much of the heat driving the seasonal melt-back of sea ice is provided from below by upwelling of warm deep water. An increased flux of freshwater (precipitation) to the Southern Ocean would lead to increased stratification and, consequently, to a reduction in the upward flux of heat. Therefore, the net effect of global warming could be an increase in sea ice cover of the Southern Ocean. The consequences for ocean ecology and carbon fluxes of this counterintuitive response to global warming remains untested.
As noted in the preceding section, wind patterns are expected to change in response to global warming, and this may lead to a domino effect with changes in upwelling, mixed layer depths, sea ice extent, storminess, etc. Each of these, in turn, can affect ecosystem structure and carbon fluxes.
Critical Questions/Hypotheses
The working group discussed a number of questions and hypotheses concerning the role of the Southern Ocean in the global carbon cycle, as well as potential responses to global warming that may impart significant feedbacks into the rise in atmospheric CO2 levels. Physical, biogeochemical and ecological issues pertaining to the Southern Ocean, jointly afforded high priority for future study, include:
General Strategy to Advance our Understanding of the Southern Ocean
Given the high degree of spatial and temporal variability in the Southern Ocean, it is necessary to extrapolate from discrete observations to obtain basin-scale estimates of fluxes and changes in ecosystem structure. Successful extrapolation will require coordinated efforts in the following areas.
Process Studies. Meaningful extrapolation requires a mechanistic understanding of the processes regulating ecosystem structure and carbon fluxes, as well as the response of these processes to changes in physical forcing. Acquisition of process-level information will require intensive dedicated studies, similar in some ways to those conducted by JGOFS, but also with some significant distinctions. For example, process studies may involve purposeful manipulations to test hypotheses concerning ecological and biogeochemical responses to physical and chemical perturbations. Process studies will also exploit natural temporal and spatial variability to explore responses to changing environmental conditions. For example, investigators may study regions "downstream" from islands (e.g., Kerguelan) to determine the long-term ecological response, as well as the net effect on nutrient consumption and export production, to a sustained supply of iron. Process studies will also be designed to exploit natural interannual variability, for example associated with the Antarctic Circumpolar Wave, to determine the sensitivity of ecosystem structure and carbon fluxes to perturbation by changes in sea ice, wind stress, and other physical boundary conditions.
Expanded Observations. While process studies enable us to establish the biogeochemical response to changes in environmental conditions, they are insufficient in themselves to extrapolate this understanding to basin and global scales. Spatial and temporal coverage of observations must be expanded through a coordinated program involving volunteer observing ships (VOS) and satellites, together with instrumented moorings and floats. Satellite data will be used to provide broad synoptic coverage, as well as time-series information, about surface properties. Volunteer observing ships provide platforms to expand the spatial and temporal coverage of parameters not measurable from space beyond results that can be obtained from dedicated research cruises. While commercial ship traffic in the Southern Ocean is rare, the frequency of visits by tourist ships is increasing, and data can be collected from research cruises dedicated to other disciplines. Moorings and floats provide additional platforms for collecting time-series information about parameters not measurable from space, as well as the opportunity for in situ calibration of satellite-derived measurements. Moored instrumentation will also provide time-series information about subsurface parameters; for example, to define the stratification of the upper water column, together with critical information about concentrations and fluxes dependent on stratification.
Models. Parameterization of the understanding derived from process studies, and incorporation of that information into regional models, should take place concurrently with the observational activities described above. High-resolution coupled models of the Southern Ocean, incorporating essential components of ecosystem structure and biogeochemistry, need to be developed. . Moreover, models with realistic thermodynamics (e.g., linking sea ice models with GCMs, convective processes on the continental shelves, etc.) also must be developed and tested. Model development should be an iterative process, with repeated tests against observational data and sequential improvements as new information becomes available. Modeling will serve the broader community, both by developing credible predictions of the response of the Southern Ocean to rising levels of atmospheric CO2 and through an improved understanding of the contribution by Southern Ocean processes to natural variability of atmospheric CO2 associated with past climate changes.