Oxygen gradients across the Pacific Ocean: Resolving an apparent discrepancy between atmospheric and ocean observations and models

We use oceanic and atmospheric model simulations to investigate and resolve a disagreement between observations of atmospheric O2/N2 and CO2 data and air-sea fluxes estimated from an ocean inversion. Atmospheric observations of O2/N2 and CO2 can be combined to calculate atmospheric potential oxygen (APO=O2/N2+1.1CO2), a powerful atmospheric tracer for ocean biogeochemical processes that is not influenced by terrestrial photosynthesis or respiration. A recent study identified a deep APO minimum in the Northwest Pacific from measurements collected on a repeat transect between New Zealand and Japan. This minimum could not be reproduced in atmospheric model simulations forced with air-sea fluxes estimated from ocean data, suggesting that oxygen uptake in the Northwest Pacific must be under-estimated by a factor of two. We use an updated ocean inverse method to estimate new air-sea fluxes from the ocean interior measurements at a higher spatial resolution than previous work using a suite of ten ocean general circulation models (OGCMs). These new air-sea flux estimates are able to match the atmospheric APO data when used as boundary conditions for an atmospheric transport model. The relative roles of thermal and biological processses in contributing to oxygen absorption by the North Pacific and other ocean regions is investigated.