Problems with using radiocarbon to infer ocean ventilation rates for past and present climates

The oceanic 14C distribution reflects mainly the circulation pattern and intensity, but is also sensitive to the exchange processes at the air–sea interface. In order to separate the relative contributions of both effects (that might have changed in the past), we incorporate in an ocean general circulation model two passive tracers, namely, the normalized radiocarbon ratio (Δ14C) and the actual age of water. We quantify, for both present and glacial conditions, the decoupling between the 14C ventilation rate and the circulation intensity as the difference between the simulated 14C age and actual age of water. The 14C age of the model Antarctic Bottom Water (AABW) appears systematically older than its actual age, the discrepancy being larger for glacial conditions because of the more extensive Antarctic sea-ice cover. Our results suggest that the AABW flow rate could have been stronger than today during the Last Glacial Maximum, contrary to what might be inferred from a naive interpretation of 14C measurements in deep-sea sediment cores.