• anthropogenic CO2;
  • bomb radiocarbon;
  • ocean modeling;
  • transient tracers

[1] The injection of radiocarbon (14C) into the atmosphere by nuclear weapons testing in the 1950s and 1960s has provided a powerful tracer to investigate ocean physical and chemical processes. While the oceanic uptake of bomb-derived 14C was primarily controlled by air-sea exchange in the early decades after the bomb spike, we demonstrate that changes in oceanic 14C are now primarily controlled by shallow-to-deep ocean exchange, i.e., the same mechanism that governs anthropogenic CO2 uptake. This is a result of accumulated bomb 14C uptake that has rapidly decreased the air-sea gradient of 14C/C (Δ14C) and shifted the main reservoir of bomb 14C from the atmosphere to the upper ocean. The air-sea Δ14C gradient, reduced further by fossil fuel dilution, is now weaker than before weapons testing in most regions. Oceanic 14C, and particularly its temporal change, can now be used to study the oceanic uptake of anthropogenic CO2. We examine observed changes in oceanic Δ14C between the WOCE/SAVE (1988–1995) and the CLIVAR (2001–2007) eras and simulations with two ocean general circulation models, the Community Climate System Model (CCSM) and the Estimating the Circulation and Climate of the Ocean Model (ECCO). Observed oceanic Δ14C and its changes between the 1980s–90s and 2000s indicate that shallow-to-deep exchange is too efficient in ECCO and too sluggish in CCSM. These findings suggest that mean global oceanic uptake of anthropogenic CO2 between 1990 and 2007 is bounded by the ECCO-based estimate of 2.3 Pg C yr−1 and the CCSM-based estimate of 1.7 Pg C yr−1.