A new look at ocean ventilation time scales and their uncertainties

Abstract A suite of eddy‐resolving ocean transient tracer model simulations are first compared to observations. Observational and model pCFC‐11 ages agree quite well, with the eddy‐resolving model adding detail. The CFC ages show that the thermocline is a barrier to interior ocean exchange with the atmosphere on time scales of 45 years, the measureable CFC transient, although there are exceptions. Next, model simulations are used to quantify effects on tracer ages of the spatial dependence of internal ocean tracer variability due to stirring from eddies and biases from nonstationarity of the atmospheric transient when there is mixing. These add to tracer age uncertainties and biases, which are large in frontal boundary regions, and small in subtropical gyre interiors. These uncertainties and biases are used to reinterpret observed temporal trends in tracer‐derived ventilation time scales taken from observations more than a decade apart, and to assess whether interpretations of changes in tracer ages being due to changes in ocean ventilation hold water. For the southern hemisphere subtropical gyres, we infer that the rate of ocean ventilation 26–27.2 σ θ increased between the mid‐1990s and the decade of the 2000s. However, between the mid‐1990s and the decade of the 2010s, there is no significant trend—perhaps except for South Atlantic. Observed age/AOU/ventilation changes are linked to a combination of natural cycles and climate change, and there is regional variability. Thus, for the future it is not clear how strong or steady in space and time ocean ventilation changes will be.