资源环境科技发展态势分析平台

The Southern Ocean (SO) carbon sink has strengthened substantially since the year 2000, following a decade of a weakening trend. However, the surface ocean pCO(2) data underlying this trend reversal are sparse, requiring a substantial amount of extrapolation to map the data. Here we use nine different pCO(2) mapping products to investigate the SO trends and their sensitivity to the mapping procedure. We find a robust temporal coherence for the entire SO, with eight of the nine products agreeing on the sign of the decadal trends, that is, a weakening CO2 sink trend in the 1990s (on average 0.22 0.24pgCyr(-1)decade(-1)), and a strengthening sink trend during the 2000s (-0.35 0.23pgCyr(-1)decade(-1)). Spatially, the multiproduct mean reveals rather uniform trends, but the confidence is limited, given the small number of statistically significant trends from the individual products, particularly during the data-sparse 1990-1999 period.

Plain Language Summary The Southern Ocean plays an important role in regulating Earth's climate as it takes up a substantial amount of carbon dioxide from the atmosphere, thereby limiting the effect of global warming. However, this part of the global ocean is also the least well observed and observational data are sparse. Therefore, to study Southern Ocean carbon uptake, data interpolation methods are used to estimate the variability of the carbon uptake from the few existing observations. This poses the question on how reliable these estimates are. The Surface Ocean CO2 Mapping intercomparison project aims to do exactly that, that is, test how reliable current estimates are by comparing results from different methods. Here we compare the results from nine data interpolation methods in the Southern Ocean from 1990 to 2010 and find a broad and encouraging agreement regarding decadal carbon uptake signals, whereas a spatially more refined analysis reveals much less agreement locally, illustrating the need to continue the measurement effort in the Southern Ocean.