To meet these goals, U.S. JGOFS has developed an approach based on large-scale surveys, time series stations, process studies, modeling, and data management. The large-scale surveys are designed to provide a basin-scale to global-scale view of biogeochemical properties on seasonal time scales. Critical properties include surface pigment, primary production CO2, and export fluxes. The time series stations will provide long-term, consistent observations to study seasonal variability of biogeochemical processes at a few sites. An improved mechanistic understanding of crucial biogeochemical processes is the objective of the process studies. These campaigns are conducted in critical regions of the ocean for a limited duration. Modeling activities will synthesize these data sets to provide a diagnostic understanding of ocean biogeochemistry as well as eventual use in predictive studies of the ocean's response to climate change. Lastly, these data sets and models will be maintained and managed so that future researchers can have confidence in the quality of the data as well as to facilitate data sharing and intercomparison.
In the area of ocean biogeochemistry, there is a dilemma of linking the smaller scale processes typically observed by biological oceanographers with the larger scale patterns measured by geochemists. In the Joint Global Ocean Flux Study (JGOFS), the time series stations at Bermuda and Hawaii (Michaels et al., 1994; Karl et al., 1996) show little coherence between the level and variability of primary productivity and downward flux of organic carbon in the mid-ocean. Part of this discrepancy may result from the different time and space scales of the processes controlling primary productivity and the vertical fluxes of organic material. Although the buffering capacity of the ocean is enormous given the large inorganic carbon pool (e.g., Siegenthaler and Sarmiento, 1993), there is considerable short-term variability in its effects on the global carbon cycle (Keeling et al., 1995).
The Southern Ocean poleward of 30°S accounts for roughly 20% of the surface area of the world ocean. Circulation is dominated by the Antarctic Circumpolar Current (ACC) which has the largest volume transport of any major current (approximately 130 Sverdrups). Much of the ventilation of the deep ocean takes place in the Southern Ocean. Through air/sea exchange and sea ice formation, upper ocean waters sink and renew deep and intermediate waters of the world ocean. Furthermore, these deep waters derive their physical, chemical, and biological characteristics through processes occurring in the upper waters of the Southern Ocean.
The strong seasonal advance and retreat of sea ice plays a critical role in the physical and biological dynamics of the Southern Ocean. The maximum extent is about 20 x 106 km2 and the minimum is about 4 x 106 km2. The ice edge behaves much like an ocean frontal system, albeit one that migrates several hundred kilometers north and south during the year. It strongly affects biological productivity as well as ocean circulation. The ice edge supports high concentrations of marine life, including higher trophic levels.
Field measurements of delta-pCO2 suggest that the Southern Ocean is a net sink of atmospheric CO2 (Takahashi et al., 1993). However, models (e.g., Tans et al., 1990) support the view that the Southern Ocean is essentially neutral with respect to the uptake and release of CO2. As the circulation is characterized by zones of divergence and convergence, we expect the patterns of uptake and release to be complex. Thus, it has proven to be extremely difficult to constrain the net CO2 flux to anything more precise than -3 to +1 GT C y-1.
Recent field work at the Polar Front (de Baar et al., 1995; Turner and Owens, 1995) reveals strong uptake of CO2, consistent with biological uptake. Increased chlorophyll and iron concentrations in the Polar Front may be linked (de Baar et al., 1995), but the source of the enriched iron concentrations is not apparent. de Baar et al. (1995) suggest that iron may be acquired as the ACC crosses shallow topographic features, but the time required for diffusion of such bottom material is too long to support the observed patterns.
The control of primary productivity in the Southern Ocean remains an enigma because of the persistent high nutrient, low chlorophyll concentrations south of the Polar Front. Various processes have been suggested from light limitation (Mitchell et al., 1991, Nelson and Smith, 1991) to iron limitation (Martin et al., 1990; Lizotte and Sullivan, 1992). These processes are not necessarily mutually exclusive. Considerable attention has been placed on trace metal limitation, and although the results are compelling, there remain some troublesome points. For example, chlorophyll concentrations increase with iron addition, but productivity is not necessarily enhanced. Bottle experiments reveal a 3-5 day lag time before the iron-enriched samples begin to deviate from the control samples. Some experiments show a strong size-selective response to iron fertilization while others do not.
The next step in understanding the Southern Ocean's role in the carbon cycle is to quantify the link between productivity and downward carbon flux. Most sediment trap programs have measured high particle flux rates only during limited time periods or in restricted locations (Honjo, 1990; DeMaster et al., 1992). The measured fluxes only represent about 10% of annual productivity. However, about 50% of the productivity is supported by nitrate ("new production"), and this is potentially available for export (Smith and Sakshaug, 1990). In other regions of the world ocean, high f-numbers are associated with high total productivity. The discrepancy between measured sediment fluxes and the apparent production that is available for export is difficult to reconcile. As with the measurements of ĘpCO2, one likely explanation is that our sampling is inadequate, and critical regions and scales are not being sampled. However, if the magnitudes of these estimates of particle flux and new production are correct, this suggests that 1) there is considerable vertical export of organic matter from the euphotic zone via pathways other than gravitational sinking or, 2) the nitrogen-based estimates of new production significantly overestimate the amount of potential production available for export.
Much as with the physical dynamics of the Antarctic Circumpolar Current, JGOFS research to date suggest that mesoscale dynamics play a more important role in regulating horizontal and vertical exchange than expected. More emphasis will be given to sampling designs that can account for these processes over long time periods.
The linkage between the physical dynamics and biogeochemical processes must be better understood if we are to make significant progress on the next challenge of climate research; how will the ocean respond to climate change? Much remains to be learned about how these two components of the ocean system operate separately, but we must design a strategy to link these two components together. The feedbacks between the physical system and the biogeochemical system are complex, and it has become apparent that mesoscale processes and intense, short-lived events play a critical role in this coupling.
The net effect of these processes is a high level of sensitivity to climate change. Changes in atmospheric forcing (winds as they affect mixing, latent heat flux as it affects sea ice formation, etc.) will have a significant impact on Southern Ocean circulation. For example, changes in the freshwater balance might be one outcome of climate change. Given that the weak stratification of the Southern Ocean is largely determined by salinity, such changes would affect the mixing regime and possibly ecosystem structure and productivity. Our ability to predict these feedbacks rests on understanding the critical processes at the appropriate time and space scales, development of numerical models that adequately represent these processes, and production of data sets with which to force and test the models.
The Southern Ocean may be viewed as a set of concentric rings that divide the region into four distinct zones. These zones have distinctive biogeochemical properties, and the zones are thought to be relatively continuous entities. The original JGOFS Implementation Plan discusses these four regions and science activities, which include