Diagnosing deglacial ocean carbon cycle change through radiocarbon and stable carbon isotopes

Radiocarbon (Δ¹⁴C) and stable carbon isotope (δ¹³C) proxy records provide important constraints on how carbon is redistributed among Earth’s surface reservoirs during major climate transitions. Previous work by Kobayashi et al. (2024, Climate of the Past) showed that transient simulations with the MIROC 4m climate model reproduce the timing of deglacial atmospheric pCO₂ changes but underestimate their magnitude. Here, we extend this analysis by using carbon isotope proxies to better diagnose ocean carbon cycle processes during the last deglaciation (21 to 11 ka BP).

We combine three-dimensional transient model output with marine sediment core and ice core records of Δ¹⁴C and δ¹³C to examine how changes in ocean ventilation, biological carbon export efficiency, and alkalinity cycling are reflected in carbon isotope budgets. Particular attention is given to Heinrich Stadial 1 (HS1), the Bølling–Allerød, and the Younger Dryas, periods characterized by abrupt changes in the Atlantic Meridional Overturning Circulation (AMOC) and pronounced interhemispheric climate asymmetry.

We analyze three-dimensional transient model output and compare the results with existing marine sediment core and ice core records of Δ¹⁴C and δ¹³C. This comparison is used to examine how changes in ocean circulation and biological carbon export and remineralization are expressed in carbon isotope budgets. We focus on Heinrich Stadial 1 (HS1), the Bolling-Allerod, and the Younger Dryas, periods associated with abrupt changes in the Atlantic Meridional Overturning Circulation (AMOC) and strong interhemispheric climate asymmetry.

The model reproduces the sequence of atmospheric pCO₂ variations across these events, but comparisons with proxy data reveal a systematic underestimation of enhanced deep-ocean ventilation during HS1, particularly in the Southern Ocean and North Pacific, as indicated by marine Δ¹⁴C records. Stable carbon isotope data further suggest that reductions in biological carbon export efficiency during HS1 are weaker in the model than implied by benthic and planktonic δ¹³C records. During the Younger Dryas, proxy records indicate a continued increase in deep-ocean δ¹³C, whereas the model simulates an opposite trend, pointing to potential biases in simulated AMOC changes, ecosystem responses, or terrestrial carbon exchange.

Overall, radiocarbon and stable carbon isotope comparisons indicate that the redistribution of carbon within the ocean is underestimated in the model. To further investigate these discrepancies, we additionally report sensitivity experiments that revisit the initialization of the Last Glacial Maximum state and assess the respective roles of ocean circulation and the biological pump in shaping deglacial carbon isotope and atmospheric pCO₂ evolution.