Ice-core records show that atmospheric CO2 rose by ~12 ppm within ~200 years at the onset of the Bølling warming about 14.7 thousand years ago — one of several abrupt centennial CO2 steps of the last deglaciation whose origin is contested. Yu et al. (Nature Geoscience, 2022) used air–sea exchange tracers reconstructed from sediment cores to propose that a sudden northward expansion of Antarctic Intermediate Water (AAIW), a mid-depth water mass inefficient at sequestering atmospheric CO2, contributed to this event; however, their own carbon-budget calculation accounts for only about 4–8 ppm, and the split between AAIW expansion, concurrent Southern Ocean outgassing, and North Atlantic Deep Water (NADW) changes remains unquantified. The problem is to produce a quantitative, uncertainty-bounded attribution of each abrupt deglacial CO2 step among these mechanisms and to test whether the same mechanism mix operated at other abrupt events such as the Younger Dryas. This attribution has not yet been established quantitatively at the centennial scale.

Background

During the last deglaciation (roughly 18–11 thousand years ago, ka), atmospheric CO2 rose about 90 ppm from its glacial level, and because the ocean holds the largest readily exchangeable carbon reservoir, this rise must reflect carbon transferred from the ocean interior to the atmosphere. The deglaciation was not smooth: Heinrich Stadial 1 (HS1, ~18–14.7 ka) was a cold interval of North Atlantic iceberg discharge, the Bølling–Allerød warm period began with an abrupt Northern Hemisphere warming at ~14.7 ka, and the Younger Dryas (YD, ~12.8–11.7 ka) was a subsequent cold relapse. High-resolution ice-core measurements from the West Antarctic Ice Sheet Divide core (Marcott et al. 2014) showed that besides the ~50 ppm millennial CO2 rise during HS1, deglacial CO2 also rose in abrupt centennial steps tied to Northern Hemisphere climate events: at the Bølling onset, CO2 rose by ~12 ppm within ~200 years, and comparable jumps occurred at other event boundaries. Which ocean processes produced these abrupt steps is contested; proposed mechanisms include reinvigoration of North Atlantic Deep Water (NADW, the deep southward-flowing limb of the Atlantic Meridional Overturning Circulation (AMOC), the basin-scale overturning that ventilates the deep ocean), outgassing from the Southern Ocean where carbon-rich deep water upwells, North Pacific processes, and permafrost carbon release. Yu et al. (2022) attacked this question with air–sea exchange tracers. From paired reconstructions of carbonate ion concentration and phosphate in benthic foraminifera (single-celled seafloor organisms whose calcium carbonate shells record the chemistry of surrounding waters), one can estimate the preformed (source-water) carbonate-ion and dissolved inorganic carbon (DIC) content of a water mass — the values it carried when it last equilibrated with the atmosphere, before in-situ remineralization altered it. Down-core changes in these tracers, denoted [CO32−]as[CO_3^{2-}]_{as} and DICasDIC_{as}, at sites bathed by a given water mass record changes in that water mass's source properties rather than local remineralization. Applied to South Atlantic intermediate-depth site GGC90 (~1.1 km) and mid-depth North Atlantic site NEAP 4K, the study inferred (i) a net loss of oceanic carbon to the atmosphere through the Atlantic sector of the Southern Ocean during HS1, corroborated by a transient Earth-system-model experiment forced with intensified Southern Hemisphere westerly winds, and (ii) an abrupt DICasDIC_{as} drop at GGC90 within ~350 years centered on the Bølling onset, interpreted as a sudden northward expansion of Antarctic Intermediate Water (AAIW, the mid-depth ~500–1200 m water mass formed in the Southern Ocean, inefficient at sequestering atmospheric CO2 because its source waters outgas CO2 before subduction). Because northward-flowing AAIW is partly entrained into NADW, an expanded, low-DICasDIC_{as} AAIW would lower the carbon-sequestration efficiency of the newly reinvigorated NADW, counteracting the CO2 drawdown otherwise expected from NADW reinvigoration. A simple calculation in the paper converts the reconstructed DICasDIC_{as} drop of 33 µmol/kg into an 18 PgC (petagrams of carbon) oceanic carbon loss, equivalent to only ~4–8 ppm of atmospheric CO2 — part of the observed ~12 ppm. The source paper states the obstacles that keep the full attribution open: spatially sparse proxy coverage of the Southern Ocean; poorly constrained instantaneous AAIW production rate and geographic extent in such budgets; unquantified basin-scale effects of AAIW entrainment on NADW's preformed carbon content; radiocarbon chronologies whose marine reservoir-age corrections (ΔR, the local offset of the ocean-atmosphere radiocarbon age difference from the global average) and bioturbation smoothing are comparable to the ~200-year event duration; a sedimentation rate at GGC90 that drops to ~3 cm/ka after ~14.3 ka, too low to resolve the Younger Dryas there; and strong seasonal and spatial heterogeneity of Southern Ocean surface pCO2 (the partial pressure of CO2 in surface seawater, which sets the direction and rate of air–sea gas exchange), which prevents extrapolating local surface proxies to basin-integrated air–sea fluxes.

Problem Statement

Quantitatively attribute the abrupt, centennial-scale rises in atmospheric CO2 during the last deglaciation — starting with the ~12 ppm rise within ~200 years at the Bølling onset (~14.7 ka) — among the candidate mechanisms: (i) expansion and property changes of Antarctic Intermediate Water, including the reduction of NADW's carbon-sequestration efficiency through entrainment of low-DICasDIC_{as} AAIW; (ii) contemporaneous Southern Ocean outgassing; and (iii) reinvigoration of NADW and the associated circulation changes. Because the AAIW term (i) already carries the effect of AAIW entrainment on NADW's carbon-sequestration efficiency, these mechanisms are not disjoint; the attribution should state the partitioning convention that assigns each shared contribution to a single mechanism and report the shares of the observed rise so that the itemized shares plus any residual for processes not itemized here (such as North Pacific processes or permafrost carbon release) reconcile to the observed rise. The attribution must carry stated uncertainties and must resolve the events at their own centennial timescale, not merely at the millennial scale. The Bølling-onset step is the mandatory target; the same question extends to the other abrupt CO2 steps of the deglaciation, tested at the end of the Younger Dryas (~11.7 ka): did the same mechanisms operate there in the same proportions, or do the abrupt events differ in their mechanism mix?

References

  1. Yu, J., Oppo, D. W., Jin, Z., Lacerra, M., Ji, X., Umling, N. E., Lund, D. C., McCave, N., Menviel, L., Shao, J. & Xu, C. (2022). Millennial and centennial CO2 release from the Southern Ocean during the last deglaciation. Nature Geoscience 15, 293–299. https://doi.org/10.1038/s41561-022-00910-9
  2. Marcott, S. A., Bauska, T. K., Buizert, C., Steig, E. J., Rosen, J. L., Cuffey, K. M., Fudge, T. J., Severinghaus, J. P., Ahn, J., Kalk, M. L., McConnell, J. R., Sowers, T., Taylor, K. C., White, J. W. C. & Brook, E. J. (2014). Centennial-scale changes in the global carbon cycle during the last deglaciation. Nature 514, 616–619. https://doi.org/10.1038/nature13799
  3. Yu, J., Anderson, R. F., Jin, Z. D., Ji, X., Thornalley, D. J. R., Wu, L., Thouveny, N., Cai, Y., Tan, L., Zhang, F., Menviel, L., Tian, J., Xie, X., Rohling, E. J. & McManus, J. F. (2023). Millennial atmospheric CO2 changes linked to ocean ventilation modes over past 150,000 years. Nature Geoscience 16, 1166–1173. https://doi.org/10.1038/s41561-023-01297-x
  4. Kobayashi, H., Oka, A., Obase, T. & Abe-Ouchi, A. (2024). Assessing transient changes in the ocean carbon cycle during the last deglaciation through carbon isotope modeling. Climate of the Past 20, 769–787. https://doi.org/10.5194/cp-20-769-2024
  5. Dai, Y. & Yu, J. (2025). Contributions of biological and physical dynamics to deglacial CO2 release from the polar Southern Ocean. Nature Communications 16, 2665. https://doi.org/10.1038/s41467-025-57677-x
  6. Karas, C., Nürnberg, D., Lambert, F., Shaffer, G. & Lamy, F. (2025). Enhanced deglacial carbon transport by Pacific southern-sourced intermediate and mode water. Nature Communications 16, 5245. https://doi.org/10.1038/s41467-025-60551-5
  7. Muglia, J., Schmittner, A., Mulitza, S., Rafter, P., Repschläger, J., Somes, C. & Wilmes, S.-B. (2025). Southern Ocean ventilation drives early deglacial ocean carbon release and CO2 rise. Research Square preprint. https://doi.org/10.21203/rs.3.rs-7491197/v1
  8. Liu, B. & Ilyina, T. (2024). Quantifying the role of ocean biogeochemistry on the deglacial atmospheric CO2 rise using transient simulations with MPI-ESM. EGU General Assembly 2024, abstract EGU24-18082. https://doi.org/10.5194/egusphere-egu24-18082
  9. Kobayashi, H., Oka, A., Obase, T., Nishida, M. & Abe-Ouchi, A. (2026). Diagnosing deglacial ocean carbon cycle change through radiocarbon and stable carbon isotopes. EGU General Assembly 2026, abstract EGU26-8803. https://doi.org/10.5194/egusphere-egu26-8803
  10. Yu, J. & Dai, Y. (2025). Millennial and centennial CO2 release from the Southern Ocean during the last deglaciation: A bottom-up approach. Goldschmidt 2025 conference abstract. https://doi.org/10.7185/gold2025.27330
  11. Dai, Y., Yu, J., Ren, H. & Ji, X. (2022). Deglacial Subantarctic CO2 outgassing driven by a weakened solubility pump. Nature Communications 13, 5193. https://doi.org/10.1038/s41467-022-32895-9
  12. Oppo, D. W., Lu, W., Huang, K.-F., Umling, N. E., Guo, W., Yu, J., Curry, W. B., Marchitto, T. M. & Wang, S. (2023). Deglacial temperature and carbonate saturation state variability in the tropical atlantic at Antarctic Intermediate Water depths. Paleoceanography and Paleoclimatology 38, e2023PA004674. https://doi.org/10.1029/2023PA004674