Changes in southern ocean ventilation inferred from repeat hydrographies: a maximum entropy approach

Abstract

This thesis investigates the ventilation of the Southern Ocean in terms of its state in the early 1990s, its decadal changes from the 1990s to the early 2000s, and the implications of these changes for anthropogenic carbon (CANT) storage. A maximum entropy (ME) approach is used to deconvolve tracer measurements to estimate the boundary propagator, which is the joint distribution of the locations and times since last ventilation. We use Southern Ocean data from five meridional repeat hydrography sections (P16, P18, A16, I8S/I9N and S3) of the World Ocean Circulation Experiment and the Climate Variability and Predictability project for CFC-11, CFC-12 temperature, salinity, oxygen, and phosphate, as well as data from the Global Ocean Data Analysis Project (GLODAP) for radiocarbon, combined with surface histories derived from atmospheric CFC-11 and CFC-12 measurements and the World Ocean Atlas and GLODAP databases. We quantify Southern Ocean ventilation in terms of the water-mass fraction older than 30 years, the fractions of water last ventilated in specific regions that tile the global ocean surface, the global and regional mean transit times, and the width-to-mean ratio of the global transit-time distribution (TTD). By independently deconvolving the original and repeat hydrographies for the boundary propagator, we find that Circumpolar Deep Water (CDW) has become older with a narrower TTD and an increased fraction last ventilated in the North Atlantic. Subantarctic Mode Water (SAMW) has become younger with a broader TTD and an increased fraction last ventilated in the Southern Ocean. Antarctic Bottom Water (AABW) has become younger. The inferred decadal changes in ventilation are consistent with increased upwelling of CDW and enhanced SAMW formation driven by strengthened mid-latitude westerly winds. The inferred boundary propagators are used to propagate CANT into the ocean interior assuming constant disequilibrium. We estimate that the section-averaged water-column CANT storage per unit area has increased by 0.56+/-0.02 mol/m2/yr in the 60S-20S latitude strip along the sections. The changes in CANT storage due to changes in ventilation correspond to a pronounced increase of CANT in SAMW. There are no significant changes in CANT for CDW due to changes in ventilation. The net effect of the changes in ventilation contributes 0.08+/-0.02 mol/m2/yr to the CANT storage in the Southern Ocean.