Constraining ventilation during deepwater formation using deep ocean measurements of the dissolved gas ratios 40 Ar/ 36 Ar, N 2 /Ar, and Kr/Ar

The concentration of inert gases and their isotopes in the deep ocean are useful as tracers of air‐sea gas exchange during deepwater formation. ΔKr/Ar, ΔN 2 /Ar, and δ 40 Ar were measured in deep profiles of samples collected in the northwest Pacific, subtropical North Pacific and tropical Atlantic oceans. For the ocean below 2000 m, we determined a mean ΔKr/Ar composition of −0.96% ± 0.16%, a mean ΔN 2 /Ar of 1.29% ± 0.21% relative to equilibrium saturation, and for δ 40 Ar a value of 1.188‰ ± 0.055‰ relative to air. These data are used to constrain high‐latitude ventilation processes in the framework of three‐box and seven‐box ocean models. For the three‐box model tracer data, we constrain the appropriate surface area of the high‐latitude region in both models to be 3.6% (+2.5%, −1.7%) of ocean surface area and the bubble air injection rate to be 22.7 (+8.8, −7.3) mol air m −2 yr −1 . Results for the seven‐box model were similar, with a high‐latitude area of 3.3% (+2.2%, −1.3%). Our results provide geochemical support for suggestions that the effective area of high‐latitude ventilation is much smaller than the region of elevated preformed nutrients and demonstrate that noble gases strongly constrain the ocean solubility pump. Reducing high‐latitude surface area weakens the CO 2 solubility pump in the box models and limits communication between the atmosphere and deep ocean. These tracers should be useful constraints on high‐latitude ventilation and the strength of the solubility pump in more complex ocean general circulation models.