Anthropogenic carbon dioxide in the South Atlantic western basin

• Published in: Journal of Marine Systems
• Main Authors: Ríos, Aida F., Vázquez Rodríguez, Marcos, Padín, X. A., Pérez, Fiz F.
• Format: Article in Journal/Newspaper
• Language: English
• Published: Elsevier 2010
• Subjects: Anthropogenic CO2; Back-calculation; Carbon storage; Water masses; South Atlantic Ocean; Austral; Western Basin; NADW; North Atlantic Deep Water; North Atlantic
• Online Access: http://hdl.handle.net/10261/48760; https://doi.org/10.1016/j.jmarsys.2010.06.010

23 páginas, 4 figuras The meridional WOCE line A17 was conducted during the austral summer of 1994 parallel to the eastern South American coast, from 55°S to 10°S, where one of the main limbs of the North Atlantic Deep Water (NADW), i.e., the southward-flowing Deep Western Boundary Current (DWBC) is found. Full-depth profiles of pH, total alkalinity and total inorganic carbon were measured and checked with analytical CO2 certified reference materials (CRMs), providing a high-quality dataset with good internal consistency for the CO2 system parameters that is well suited for anthropogenic CO2 (CANT) estimation. For the first time in the westernAtlantic basin the CANT has been calculated using four independent approaches and results are compared. The methods considered are the CFC-based TTD method and the φCT°, TrOCA and ∆C* carbon system-based back-calculation methods. All four methods have produced CANT distribution patterns that are in general good agreement: maximum concentrations of CANT (50–60 μmol kg− 1) are predicted for the upper warm SouthAtlantic central waters from the tropical gyres, while the minima (~ 5 μmol kg− 1) are located in the old northward-flowing branch of Circumpolar Deep Water. There are, however, some discrepancies detected. The TrOCA method yields the highest overall [CANT] values, even over the theoretical limit of CANT saturation for 1994 in the upper layers. The ∆C* approach consistently yielded negative estimates of CANT below 2800 dbar, even after correcting a reported − 8 μmol kg− 1 bias in the alkalinity measurements of the WOCE A17 line. The main overall difference between the four methods corresponds to the relative CANT maximum associated with the lower limb of NADW: this structure is well identified in the φCT° and TTD methods but seems to disappear in the case of TrOCA and ∆C*. In agreement with other intercomparison studies of CANT, the specific inventories are significantly higher (~ 45%) than those reported in the GLODAP database obtained from the ΔC* method. This ...