A float-based autonomous three-year time-series of ODZ biogeochemistry in the Eastern Tropical N. Pacific

Open ocean oxygen deficient zones (ODZ’s) host unique subsurface biogeochemical processes that have global impacts including fixed nitrogen loss and the cycling of N₂O. They have been predicted to expand geographically in response to global warming though contrary perspectives are available. In addition, biogeochemical activity is highly variable in time and space as associated with coastal upwelling plumes and certain types of mesoscale eddies. Correspondingly, ship-based observations are insufficient to both capture relevant scales of variability and provide the sustained time series records required to detect long term (≥ decadal) changes. Sustained autonomous observation using float platforms addresses these problems but the unique subsurface biogeochemistry of ODZ’s requires sensor capabilities that up to now have not been available. In the cores of ODZ’s, O₂ is often undetectable even with nM sensitivity and key ODZ microbial nitrogen cycle processes now appear to have O₂ sensitivities in the 0 to 3 μM range. The buildup of biogenic N₂ quantitatively measures N-loss in ODZ’s but is analytically challenging particularly for autonomous deployment. We have deployed an Argo-style float in the Eastern Tropical N. Pacific ODZ with sensors designed to measure both nM-level O₂and N₂ gas concentration at the levels of precision and stability required for sustained study of ODZ biogeochemistry. The ODZ-Float has operated continuously for the last three years within the ETNP producing a data set unique with respect to its resolution and longevity. During this time O₂ below the oxycline has was always <100 nM even though float trajectory would suggest crossing of the northern ODZ boundary according to the WOA18 database. Biogenic N₂ was measured with a precision of ± 0.15 µmol/kg and showed apparent seasonal variations. During 2-week isopycnal drifts, in situ rates of N₂ production were calculated for the first time that showed large temporal variations.