ATMOSPHERIC SCIENCES

Keeping Our Cool: Does the Ocean Dampen the Greenhouse Effect?

    As the burning of fossil fuels in the last century raised atmospheric carbon dioxide levels, was the ocean a source or a sink for CO₂—or both? A comprehensive survey of CO₂ in the world's oceans may give scientists the chance to find out.

By Christopher L. Sabine, Princeton University; Douglas W. R. Wallace, Brookhaven National Laboratory, Upton, N.Y.; and Frank J. Millero, University of Miami, Miami, Fla.

Industrial emissions of carbon dioxide (CO₂, the chief greenhouse gas), have more than tripled in the past 40 years, causing concern that increased heat-trapping gas in the atmosphere could cause global warming. Predicting long-term climate trends is tricky, particularly if the capacity of the ocean to take up or give off CO₂ is unknown. Estimates based on model results suggest that the oceans could remove approximately 37% of the CO₂ produced each year from the burning of fossil fuels, but to date, scientists have not been able to confirm this number with direct oceanographic measurements. It is important to know how much CO₂ the oceans are taking up because they may help to reduce the amount of warming caused by the greenhouse effect.

    In 1979, a paper published in Science pointed out that model estimates of CO₂ uptake could only be confirmed with oceanic CO₂ measurements made at accuracy levels that were not technologically possible at that time. Seventeen years later, the technology has improved, and a global survey of CO₂ is nearly complete. Enough high quality data have finally been gathered to determine the distribution and mass of anthropogenic CO₂ stored in the ocean.

    January 1996 marked the completion of one ambitious part of the global oceanic inorganic carbon survey: a 14-month, 92,000-km-long research cruise in the Indian Ocean. This survey, conducted with the international World Ocean Circulation Experiment- Hydrographic Program (WOCE-HP), is part of a project involving 10 U.S. universities and national laboratories that have been working together since 1990 to acquire a high-quality inorganic carbon data set for all of the world's oceans.

    Why are so many accurate carbon measurements in the oceans needed?

A Small Change in a Big Number

The CO₂ measurements must be highly accurate because scientists are trying to quantify a very small change in a very big number. Compared to the tremendous amount of carbon naturally stored in the ocean, the increase in CO₂ from human activities is very small. Approximately 98.5% of the CO₂ in the ocean-atmosphere inventory is found in the ocean. The oceans hold most of the carbon because CO₂ is approximately 30 times more soluble in water than other common gases. CO₂ molecules also react with water in the ocean to form carbonic acid and its dissociation products: carbonate and bicarbonate ions.

    Because the oceans naturally contain much more carbon than the atmosphere, a dramatic industry-generated increase in CO₂ concentration—such as that observed in the atmosphere—is not seen in the oceans. The anticipated total CO₂ (TCO₂) increase in surface water is approximately 0.05% per year—only one-tenth of the increase observed in the atmosphere. Detecting a signal that small, given the complicated carbon chemistry in the ocean, requires extremely accurate measurements.

Global Variations in Oceanic CO₂

    The figures illustrate the horizontal and vertical changes in oceanic CO₂ levels observed in the Indian Ocean. Fluctuations in these CO₂ levels in the upper ocean can be 10 to 100 times more than the annual anthropogenic increase. Therefore researchers have to make hundreds of thousands of CO₂ measurements to accurately distinguish the natural variability from the CO₂ increase due to rising atmospheric concentrations.

Fig. 1. Maps of a) the distribution of total CO₂ and b) the distribution of total alkalinity on a constant density surface ([theta-sigma] = 27.5; depth range 1100–1400 m) in the Indian Ocean. There were 1244 stations occupied as part of the WOCE Indian Ocean Survey (locations identified with black dots on the map)—10 times the number of Geochemical Ocean Sections Survey sampling sites 18 years earlier.

    Uptake of carbon in the surface water by marine organisms for photosynthesis and the subsequent remineralization of the organic matter at depth after the organisms die strongly influence the distribution of carbon in the oceans. The fact that CO₂ is more soluble in cold, polar water than in warm water also influences the distribution. These two processes, (termed the biological and solubility pumps), and the relatively slow mixing time of the oceans, result in large horizontal and vertical variations of CO₂ in the oceans. (see figures)

    The maps along the [theta-sigma] = 27.5 density surface in the figure show that both the TCO₂ and total alkalinity (TA) concentrations increase from south to north in the Indian Ocean. The net water flow along this surface is from south to north, so the northern waters have had more time to accumulate carbon from the remineralization of organisms sinking from the surface. The concentrations of both TCO₂ and TA are generally higher in the eastern basin than at the same latitude in the west, reflecting the contribution of relatively low-carbon Atlantic water from around the southern tip of Africa.

    The highest CO₂ concentrations are observed in the deep waters at the northern end of both sections (see figure). The lowest concentrations occur in northern surface waters where the biological pump removes carbon. Beyond these similarities, the TCO₂ and TA distributions in the upper waters are very different. For example, the Antarctic Intermediate Water can be observed as a TA minimum where it leaves the surface at approximately 50°S (the blue-green tongue between the yellow regions), but it is not readily apparent in the TCO₂ section.

    The difference in the vertical gradients of TCO₂ and TA shows that each of the carbon parameters is influenced differently by the various processes. For example, TCO₂ is influenced by photosynthesis and remineralization to a much greater degree than TA. The constant rain of organic matter from the surface waters and subsequent remineralization maintain the sharp increase in TCO₂ with depth that masks the Antarctic Intermediate Water signal. By comparing and contrasting the distributions of the different carbon parameters, the processes that transport carbon within the ocean can be evaluated—an important first step in estimating the mass of anthropogenic CO₂ stored in the ocean

Techniques for Studying Anthropogenic CO₂ in the Ocean

Three approaches, which were limited in the past due to both the quantity and quality of the available data, are now being employed with the survey data to quantify the oceanic sink for anthropogenic CO₂. The first approach uses measurements of the difference in the sea and air CO₂ levels together with wind speed observations to calculate the flux of CO₂ across the sea-air interface. Scientists can determine how much CO₂ is entering the ocean by summing up this flux over the entire ocean. This approach is the most direct way to estimate the oceanic CO₂ uptake, but it is severely limited by spatial and temporal variability in the oceans. By combining recently obtained measurements of pCO₂ (partial pressure of CO₂ gas) with the existing data set, these uncertainties can be reduced.

    The differences of pCO₂ in ocean water and in air along the cruise tracks in the Indian Ocean demonstrate the wide spatial differences that cause the ocean to be a sink or a source of CO₂, depending upon the location and conditions (see figure). Positive values indicate that CO₂ is moving from the ocean to the atmosphere, and negative values indicate that the oceans are taking up CO₂. The high sea water CO₂ concentrations relative to the atmosphere in the Arabian Sea region (red, yellow, and green) indicate that this area was a potential significant source of CO₂ to the atmosphere in the summer of 1995, while the southwestern Indian Ocean was a sink for CO₂ at the time of sampling.

Fig. 2. Map of surface [Delta]pCO₂ (sea-air)

    The high seawater values in the Arabian Sea were caused by the upwelling of CO₂-rich waters associated with the southwest monsoon, a natural phenomenon with strong seasonality. This example indicates why we need to understand naturally occurring variations before quantifying the anthropogenic contributions.

    A second approach is to evaluate the increase in carbon in the interior of the ocean due to the invasion of anthropogenic CO₂. One "time-series" method uses measurements of TCO₂ made at a certain time to develop a predictive equation based on the relationship between the observed TCO₂ and salinity, temperature, oxygen and total alkalinity (or silicate), which are measured at the same time. These relationships hold over large spatial scales, and their use drastically reduces the complicating effects of natural variability in surface waters by examining the signal in the more stable, deeper waters.

    The difference between TCO₂ measurements made at different times (for example over decades) will isolate the anthropogenic contribution. Systematic changes in the difference between the predicted and measured TCO₂ concentrations over time provide a direct estimate of oceanic CO₂ flux caused by the uptake of anthropogenic CO₂.

    This technique was applied to the recently collected WOCE Indian Ocean survey data set and data from the 1977–1978 Geochemical Ocean Sections Survey (GEOSECS) of the Indian Ocean. The difference between the measured and predicted TCO₂ concentrations for the same area of the Indian Ocean examined 18 years apart is consistent with the anthropogenic CO₂ uptake anticipated to be ~1 µmol kg^-1 yr^-1 for the 18 years since GEOSECS.

    Another method of estimating the anthropogenic CO₂ inventory uses the difference ([Delta]C*) between the WOCE Indian Ocean TCO₂ concentrations (corrected for the influence of the biological pump) and the TCO₂ concentration these waters would have had when they were formed at the surface assuming a preindustrial atmospheric CO₂ concentration of 280 µatm. Since the deep waters (&#;2000 m) actually were formed before anthropogenic CO₂ was added to the atmosphere, they provide a baseline for the calculations. The difference between [Delta]C* in deep (&#;2000 m) water and in shallower waters can be interpreted as the anthropogenic inventory. This inventory is larger than the "time-series" inventory because it integrates changes since preindustrial time, not just since the last 18 years. These assessments imply that approximately 28% of the total anthropogenic CO₂ inventory in this area of the Indian Ocean has accumulated in the last 18 years.

    The distribution of CO₂ between sea and air has changed dramatically during glacial and interglacial climate changes. A critical issue in predicting future atmospheric CO₂ levels, and in developing policies designed to control CO₂ increases in the future, is how the oceanic uptake might change as atmospheric CO₂ levels continue to increase and if the climate changes.

    Direct measurements help to determine the current or past oceanic responses to anthropogenic CO₂. To predict how the oceans will respond in the future, however, requires the use of three-dimensional global carbon models. An important test of how these models perform lies in how well they can reproduce current ocean carbon distributions and transports. The global CO₂ survey is finally providing the comprehensive data set necessary for this model validation.

    The cooperative research of numerous scientists has generated a quantity and quality of data far greater than previously achieved. Data from the current global survey are being made available on a cruise by cruise basis. As these data become finalized, they will be archived and made available to the public with an array of past carbon work at the Carbon Dioxide Information Analysis Center at the Oak Ridge National Laboratory. The only major region that remains to be sampled is the North Atlantic Ocean, which was last sampled extensively during 1981–1983. The resulting data set will be a resource used for decades by scientists investigating regional and global-scale carbon cycle problems and those involved in making decisions about energy policy and future climate.

    For more information on the Global CO₂ survey and the availability of carbon data, refer to the World Wide Web site: http://cdiac.esd.ornl.gov/cdiac/oceans/home.html.

    Source: Eos, February 4, 1997, p. 49.

GLOSSARY

anthropogenic CO₂—Carbon dioxide released to the atmosphere by human activities;