Anthropogenic CO2 uptake by the ocean based on the global chlorofluorocarbon data set

Science

Volumn 299, Issue 5604, 2003, Pages 235-239

  McNeil, Ben I ^a   Matear, Richard J ^b   Key, Robert M ^a   Bullister, John L ^c   Sarmiento, Jorge L ^a  

^a PRINCETON UNIVERSITY   (United States)

^b ANTARCTIC CLIMATE AND ECOSYSTEMS COOPERATIVE RESEARCH CENTRE   (Australia)

^c National Oceanic/Atmospheric Admin   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

CHLOROFLUOROCARBONS; OCEANOGRAPHY;

OCEANIC INVENTORY;

CARBON DIOXIDE;

CARBON DIOXIDE; CHLOROFLUOROCARBON; DISSOLVED ORGANIC MATTER;

AIR-SEA INTERACTION; ANTHROPOGENIC EFFECT; CARBON CYCLE; CARBON DIOXIDE; CARBON FLUX; GENERAL CIRCULATION MODEL; GLOBAL OCEAN;

ANALYTIC METHOD; ARTICLE; ATMOSPHERE; CALCULATION; CARBON DIOXIDE MEASUREMENT; ENVIRONMENTAL PROTECTION; GLOBAL CHANGE; MATHEMATICAL MODEL; MEASUREMENT; OCEAN ENVIRONMENT; PRIORITY JOURNAL; SOLUBILITY;

EID: 0347926185     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.1077429     Document Type: Article

References (41)

1. R. J. Matear, A. C. Hirst, B. I. McNeil, Geochem. Geophys. Geosys. 1, 1 (2000).
2. K. Keller, R. Slater, M. Bender, R. M. Key, Deep-Sea Res. 2 49, 345 (2002).
3. L. Bopp et al., Global Biogeochem. Cyc. 16, 1 (2002).
4. R. Keeling, H. Garcia, Proc. Natl. Acad. Sci. U.S.A. 99, 7848 (2002).
5. G.-K. Plattner, F. Joos, T. F. Stocker, Global Biogeochem. Cycl. 16, 4 (2002).
6. R. F. Keeling, S. C. Piper, M. Heimann, Nature 381, 218 (1996).
7. M. Battle et al., Science 287, 2467 (2000).
8. J. T. Houghton et al., Climate Change 2001: The Scientific Basis, Contribution of Working Group 1 to the Third Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge Univ. Press, Cambridge, 2001).
9. J. C. Orr et al., Global Biogeochem. Cyc. 15, 43 (2001).
10. T. Peng et al., Nature 396, 560 (1998).
11. D. W. R. Wallace, Monitoring Global Ocean Carbon Inventories [Ocean Observing System Development Panel (OOSDP) Background Report 5, Texas A&M Univ., College Station, TX, 1995].
12. C. L. Sabine et al. Global Biogeochem Cyc. 13, 179 (1999).
13. B. I. McNeil, B. Tilbrook, R. J. Matear, J. Geophys. Res. 106, 31431 (2001).
14. M. H. England, E. Maier-Reimer, Rev. Geophys. 39, 29 (2001).
15. 2 can be dissolved in the surface ocean [E. T. Sundquist, L. N. Plummer, T. M. L. Wigby, Science 204, 1203 (1979)].
16. R. J. Matear, unpublished data.
17. Y. W. Watanabe, T. Ono, A. Shmamoto, Mar. Chem. 72, 297 (2000).
18. M. J. Warner, R. F. Weiss. Deep-Sea Res. 2 32, 1485 (1985).
19. S. J. Walker, R. F. Weiss, P. K. Salameh. J. Geophys. Res. 105, 14285 (2000).
20. -6 atmospheres and α is the temperaturedependent solubility (18). The time since last contact with the atmosphere (water age) was then estimated by finding the date at which the atmospheric pCFC-12 equaled the oceanic pCFC-12. Unlike CFC-11, CFC-12 continued to increase in the atmosphere through the mid-1990s and therefore will give less bias due to mixing (23). Mixing beyond bulk advection from a single source region is not accounted for when calculating CFC ages. The biases in using CFC ages were explored with an OGCM (34).
21. T. M. Hall, T. W. N. Haine. J. Phys. Oceanogr. 32, 1987 (2002).
22. D. H. Min, J. L. Bullister, R. F. Weiss. Geophys. Res. Lett. 27, 3909 (2001).
23. R. Sonnerup, Geophys. Res. Lett. 28, 1739 (2001).
24. T. W. N. Haine, T. M. Hall, J. Phys. Oceanogr. 32, 1932 (2002).
25. R. J. Matear, B. I. McNeil, unpublished data.
26. C. D. Keeling, T. P. Whorf, Trends: A Compendium of Data on Global Change (Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN, 2002).
27. C. Goyet, P. G. Poisson, Deep-Sea Res. 36, 2635 (1989).
28. We used the dissociation constants derived from (27) to calculate the equilibrium DIC concentration. Using different constants resulted in negligible changes in our estimates.
29. The preformed alkalinity distribution is essential because it largely determines the oceanic buffer capacity. Preformed alkalinity was derived using a multiple linear regression of the surface (<60 dbar) alkalinity distribution for each of the three ocean basins in a way similar to the technique used by Sabine et at. (12). Using in situ alkalinity instead of preformed alkalinity resulted in negligible changes in our results.
30. 2 is assumed to be a perturbation propagating through the ocean as a tracer inert to the biological carbon pump.
31. 2 on shorter time scales (such as decadal) during which the disequilibrium is closer to constant than over longer time scales (since preindustrial times).
32. R. J. Matear, A. C. Hirst, Tellus 51B, 722 (1999).
33. 2 from 1980 to 1989 and from 1990 to 1999 was calculated on a 1° grid at 100 m intervals and integrated to 5000 m for the Atlantic and 2000 m for the Indian and Pacific Oceans. The objective mapping techniques used were based on the methods described by Sarmiento et at. (38). For the basinwide inventories, the Southern Ocean was split from 120°E to 70°W for the Pacific, 70°W to 30°E for the Atlantic, and 30°E to 120°E for the Indian Ocean. Aside from the Southern Ocean, the major continents were taken to be main boundaries separating the ocean basins.
34. 2 inventory from 1980 to 1999 (32.8 Pg of C). We are hesitant to apply this "correction" because there is good agreement between our CFC age-based inventory and those using the direct techniques (MLR and isopycnal) in the Indian Ocean (Fig. 1), which suggests that the model bias may be exaggerated. Instead, we expand our estimated uncertainty to ∼15%, which includes both the predicted "bias" from our model analysis (including mixing and disequilibrium) and the uncertainty in the inventories obtained from the MLR and isopycnal analyses. The measurement precision of CFC was taken to be ±0.005 pmol/kg. Adding this value to the simulated CFCs in the model resulted in an uncertainty of about 5% in the global inventory. Repeat measurements of CFCs show differences in CFC ages from year to year (39). As model simulations may underestimate interannual variability in the ocean, we added extra "noise" to the CFC measurements in the model (±0.01 pmol/kg), which resulted in a further 10% uncertainty in our estimates. The uncertainty associated with the inventory calculation was determined using two different interpolation schemes (objective and loess) and was estimated to be 2%. The total uncertainty of our estimate (∼20%) was calculated by a sum-of-squares propagation of all uncertainties with the assumption that each is independent of the others. We believe the uncertainty of our estimates to be an upper limit, given the good consistency between our estimated inventory in the Indian Ocean and those using the direct methods (10, 12) (Fig. 1). Because these direct techniques (MLR and isopycna[) are not subject to the uncertainties associated with the CFC age method (for example, in ages and disequilibrium), they provide evidence of upper-bound errors.
35. The predicted warming of the ocean in the climate change simulation (0.3°C) was in good agreement with observations from 1950 to 1998 (40) and provides confidence that the model is simulating recent changes in the ocean.
36. 2 uptake, models use constant climate and assume that biological processes are operating in a steady state. These are the same assumptions used for the CFC age technique.
37. J. C. Orr et al., paper presented at the International Geosphere-Biosphere Programme Open Science Conference, Amsterdam, 10 to 13 July 2001. For more details and to download the results of this study, go to www.ipsl.jussieu.fr/OCMIP/.
38. J. L. Sarmiento, J. Willebrand, S. Hellerman, Objective Analysis of Tritium Observations in the Atlantic Ocean During 1971-1974 (Ocean Tracer Laboratory Tech Report 1, Princeton Univ. Press, Princeton, NJ, 1982).
39. S. C. Doney, J. L. Bullister, R. Wanninkhof, Geophys. Res. Lett., 25, 1399 (1998).
40. S. Levitus, J. I. Antonov, T. P. Boyer, C. Stephens, Science 287, 2225 (2000).
41. This study would not have been possible without the efforts of those responsible for collecting and analyzing CFCs during WOCE and making those measurements available. They include R. Fine, M. Rhein, W. Roether, W. Smethie, M. Warner, Y. Watanabe, R. Weiss, and C. S. Wong. We also thank three anonymous reviewers and C. Sweeney for comments on the text. B.I.B. and R.M.K. were supported by NSF (grants OCE9819144 and OCE9986310), J.L.B. was supported by NOAA's Office of Global Programs and Global Carbon Cycle Program, J.L.S. was supported by the Carbon Modeling Consortium, and R.J.M. was supported by Environment Australia and the CSIRO Climate Change Program.