Decadal accumulation of anthropogenic CO2 in the Southern Ocean: A comparison of CFC‐age derived estimates to multiple‐linear regression estimates

Abstract

The decadal accumulation of anthropogenic CO2 from 1968 to 1996 was estimated from CFC‐ages along three WOCE sections (SR3, P14, P15) in the Southern Ocean. We explore the validity of the method in the Southern Ocean by directly comparing the CFC‐age derived estimates to values obtained using a multiparametric linear regression method over the same period. The agreement in anthropogenic CO2 accumulation between the two independent methods was very good (less than ±8%) for waters younger than 30 years. We found waters older than 30 years to significantly overestimate the decadal accumulation when using the CFC‐age method. This is mainly due to significant mixing biases when using CFCs to infer ages for these older waters [Sonnerup, 2001]. The CFC‐age method assumes a steady state circulation. CFC‐11 variations on interannual timescales have been found to be relatively large [Doney et al., 1998]. To test the biases from interannual variations, we compare CFC data on the SR3 section for two different years (1991 and 1994). Although we found large differences in the estimated CFC‐11 ages (∼8 years) from the two cruises, these differences had little impact on the decadal accumulation of anthropogenic CO2 estimated by the CFC‐age method. We found the CFC‐age method to be insensitive to interannual variations or non‐ideal “age” biases [Sonnerup, 2001; Haine and Hall, 2002] because of the similar atmospheric histories of CFC‐11 and CO2 from about 1965 to 1993. Our study validates the use of transient tracer ages to infer decadal anthropogenic CO2 accumulations in the Southern Ocean during the 1970s, 1980s, and 1990s. The application of other transient tracers to determine ages will be needed to improve the estimates of anthropogenic CO2 accumulation and uptake by the Southern Ocean for both now and into the future.