The deep Atlantic breathes through one small sea, and nobody was metering it

Cornell researchers report in Nature Geoscience, published Aug. 17, that the oxygen sustaining deep-sea life across the North Atlantic comes from the churning waters of the Labrador Sea ^[1]. That matters because earlier work had concluded the Labrador Sea contributes little to the strength of the Atlantic Meridional Overturning Circulation, the variable that deoxygenation risk work has largely tracked, while this study finds it plays a critical role in oxygen transport ^[2].

The physical argument is not exotic. Ocean layers of differing temperature and density mostly refuse to mix, and co-author Palter describes the Atlantic as having a lid on it, so oxygen entering from the air stays largely in the surface layer ^[3]. There is no photosynthesis below a certain depth, so according to first author Una Miller the only atmospheric oxygen in the deep ocean arrives via overturning circulation, as surface waters that flowed through the Labrador Sea cool, densify and sink ^[4]. In the Labrador Sea, between Greenland and Newfoundland, AMOC waters turn in a gyre and oxygen-rich surface water mixes downward ^[5]. It is one of very few places where that mixing happens ^[6].

The number is the point. The team quantified the export at more than 27 teramoles of oxygen per year ^[7], which at the molar mass of O2 works out to roughly 860 million tonnes a year ^[8]. That figure matches independent estimates of respiration rates for microbes and animals across the North Atlantic deep sea ^[9]. Palter's framing is that supply from these processes balances consumption over pretty much the whole deep North Atlantic, and that animals suffer once oxygen dips below a threshold ^[10]. Miller's operational conclusion: if you want to understand the future under deoxygenation trends, AMOC strength alone will not do, and you have to understand Labrador Sea processes ^[11].

The measurement history explains why this was not known sooner. Miller's team used 60 oxygen sensors attached, for the first time, to moorings along the bottom of the Labrador and western Irminger seas ^[12]. Palter says nobody had successfully sustained multiple years of oxygen measurements on moorings like these, and that a machine learning method was needed to fill gaps so the values could be mapped ^[13]. The sensors went in during 2020 and sat for two years with no guarantee they would survive, followed by a multiyear effort to make the messy data usable ^[14].

Context for the timing: AMOC has weakened over the past 75 years and its vulnerability is contested, with warnings that a collapse would disrupt weather and damage ecosystems ^[15]. Global ocean oxygen is declining as temperatures rise, and the authors say their work sheds light on processes that may be helping the North Atlantic hold its oxygen levels ^[16].

What to watch: whether oxygen sensors become standard on subpolar moorings rather than a one-off 2020 deployment ^[12][13], and whether deoxygenation projections start resolving Labrador Sea convection as a separate term instead of scaling oxygen supply off overturning strength ^[2][11]. Also watch the reconciliation work, since a supply estimate that matches demand estimates across a basin is a result that invites independent recomputation ^[9].