Ocean carbonate system variability in the North Atlantic Subpolar surface water (1993–2017)

Abstract

The North Atlantic is one of the major sinks for anthropogenic CO2. In this study, we investigate the evolution of CO2 uptake and ocean acidification in the North Atlantic Subpolar Gyre (50°N-64°N) using repeated observations collected over the last three decades in the framework of the long-term monitoring program SURATLANT (SURveillance de l'ATLANTique). Data obtained between 1993 and 1997 suggest an important reduction in the capacity of the ocean to absorb CO2 from the atmosphere during summer, due to a rapid increase in the fugacity of CO2 (fCO2) in surface waters (5 times 15 faster than the increase in the atmosphere). This was associated with a rapid decrease in surface pH (of the order of -0.014/yr) and was mainly driven by a significant warming and increase in DIC. Similar trends are observed between 2001 and 2007 during both summer and winter with a mean decrease of pH between -0.006/yr and -0.013/yr. These rapid trends are mainly explained by a significant warming of surface waters, a decrease in alkalinity during summer and an increase in DIC during winter. On the contrary, data obtained during the last decade (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017) show a stagnation of surface fCO2 (increasing the 20 ocean sink for CO2) and pH. These recent trends are explained by the cooling of surface waters, a small decrease of total alkalinity and the near-stagnation of dissolved inorganic carbon. Overall our results show that the uptake of CO2 and ocean acidification in the North Atlantic Subpolar Gyre is substantially impacted by multi-decadal variability, in addition to the accumulation of anthropogenic CO2. As a consequence, the future evolution of air-sea CO2 fluxes, pH and the saturation state of surface waters with regards to aragonite and calcite remain highly uncertain in this region. 25

Contrasting drivers and trends of ocean acidification in the subarctic Atlantic

“…Located outside the main ocean currents of the NA, both stations similarly undergo the warming and longterm salinification observed during the last decades 23,43,46,47 . The IRM-TS shows a long-term surface warming of 0.063 ± 0.005 (Table 2), which is consistent with those shown recently by Leseurre et al 25 (0.05 ± 0.01 C yr −1 between 1993 and 2017) in the northern Irminger Sea. The lower warming trend shown by these authors is due to the recent cooling period between 2008 and 2017 (− 0.05 ± 0.01 per year), which was also observed by Frob et al 27 .…”

“…The acidification trends found here in the surface layer at the IS-TS and IRM-TS are within the values observed in other time series collected by Bates and others 6 . The trend of pH T decrease at the IRM-TS (1.70 ± 0.19 10 −3 yr −1 ) is very close to that determined by Leseurre et al 25 in the northern Irminger Sea (1.6 ± 0.1 10 −3 yr −1 using winter data) and to that determined by Garcia-Ibañez et al 14 in the same basin from hydrographic sections of the A-25 GOSHIP line from 1991 to 2015 (1.8 ± 0.1 10 −3 yr −1 ), while Vázquez-Rodríguez et al 11 obtained the same trend between 1981 and 2008 for the same section. Lauvset et al 50 found a similar trend of 2.0 ± 0.4 10 −3 yr −1 for the North Atlantic subpolar seasonally stratified (NA-SPSS) biome from 1991 to 2011.…”

“…These trends are significantly larger than those reported for the same period by Bates et al 6 (1.4 ± 0.5 10 −3 yr −1 ) associated with an increase in DIC of 1.22 ± 0.27 µmol kg −1 yr −1 . Apparently, an exchange of those Iceland and Irminger trends reported by Bates et al 6 in their Table 2 would be much more comparable with the trends described here, in addition to those obtained by Olafsson et al 9 , Vázquez-Rodríguez et al 11 , Lauvset et al 50 , Garcia-Ibañez et al 14 and Leseurre et al 25 . In the subarctic western North Pacific, the winter mixing layer (K2 and KNOT time series stations) showed significantly lower acidification trends (0.80 ± 0.4 10 −3 yr −1 ) for the 1997-2011 12 period, which was attributed to the damping generated by a strong increasing trend in TA.…”

“…The lower warming trend shown by these authors is due to the recent cooling period between 2008 and 2017 (− 0.05 ± 0.01 per year), which was also observed by Frob et al 27 . In the same region and time period, a similar long-term trend of salinification (0.005 ± 0.001 yr −1 ) was also observed by Leseurre et al 25 , which is consistent with interannual freshening episodes (Fig. 2).…”

“…The extreme weather conditions of subarctic regions make it difficult to achieve long time series with enough frequency to characterize the seasonal cycle and, consequently, to determine the long-term trends and interannual variability. Several studies have observed divergent trends with the expected increase in CO 2 in the atmosphere [25][26][27] depending on the time frame of the observations. In the subpolar NA, McKinley et al 28 find that the long-term trends in surface pCO 2 take 25 years to emerge from changes due to variability on a decadal scale and converge with the atmospheric trend.…”