Estimating the storage of anthropogenic carbon in the subtropical Indian Ocean: a comparison of five different approaches

• Main Authors: Alvarez, M., Lo Monaco, C., Tanhua, T., Yool, A., Oschlies, A., Bullister, J. L., Goyet, C., Metzl, N., Touratier, F., Mcdonagh, E., Bryden, H. L.
• Format: Text
• Language: English
• Published: Copernicus GmbH 2009
• Subjects: envir; geo; Antarctic; Indian; The Antarctic; Antarc*
• Online Access: https://doi.org/10.5194/bgd-6-729-2009; https://archimer.ifremer.fr/doc/00293/40431/38985.pdf; https://archimer.ifremer.fr/doc/00293/40431/

The subtropical Indian Ocean along 32° S was for the first time simultaneously sampled in 2002 for inorganic carbon and transient tracers. The vertical distribution and inventory of anthropogenic carbon (CANT) from five different methods: four data-base methods (ΔC*, TrOCA, TTD and C0IPSL and a simulation from the OCCAM model are compared and discussed along with the observed CFC-12 and CCl4 distributions. In the surface layer, where carbon-based methods are uncertain, TTD and OCCAM yield the same result (7±0.2 mol C m−2), helping to specify the surface CANT inventory. Below the mixed-layer, the comparison suggests that CANT penetrates deeper and more uniformly into the Antarctic Intermediate Water layer limit than estimated from the ΔC* method. Additionally, significant CFC-12 and CCl4 values are detected in bottom waters, associated with Antarctic Bottom Water. In this layer, except for ΔC* and OCCAM, the other methods detect significant CANT values. Consequently, the lowest inventory is calculated using the ΔC* method (24±2 mol C m−2) or OCCAM (24.4±2.8 mol C m−2) while TrOCA, TTD, and C0IPSL lead to higher inventories (28.1±2.2, 28.9±2.3 and 30.8±2.5 mol C m−2, respectively). Overall and despite the uncertainties each method is evaluated using its relationship with tracers and the knowledge about water masses in the subtropical Indian Ocean. Along 32° S our best estimate for the mean CANT specific inventory is 28±2 mol C m−2. Comparison exercises for data-based CANT methods along with time-series or repeat sections analysis should help to identify strengths and caveats in the CANT methods and to better constrain model simulations.