Model-data comparison of Atlantic delta 18O of seawater gradient changes between the LGM and Holocene

The MARGO compilation of planktonic 𝛿 ¹⁸O suggests a change of 0.37+/-0.29 ‰ in the 𝛿 ¹⁸O_sw gradient between the low and high latitude surface Atlantic from the LGM to the Holocene (Waelbroeck et al., 2014). This 𝛿 ¹⁸O_sw gradient change likely results from the combined effects of many changes in ocean circulation and atmospheric water vapor transport. To identify the relative importance of different drivers of 𝛿 ¹⁸O _sw gradient change, we constructed a 3-box model of the Atlantic with temperature-dependent Rayleigh fractionation and atmospheric water vapor transport changes derived from PMIP3 model runs (Hourdin et al., 2013; Notz et al., 2013; Schmidt et al., 2014). Using realistic ocean circulation parameters, this model closely matches modern-day 𝛿 ¹⁸O _sw values (LeGrande and Schmidt, 2006). Because of the large number of adjustable model parameters, there are multiple ways to fit the observed MARGO gradient change for the LGM. For example, a good fit is achieved with a 40% reduction in overturning, 37% less evaporation and latitudinal water vapor transport, and 75% less Arctic runoff at an 𝛿¹⁸O value of approximately -40‰. Sensitivity tests of LGM parameter changes demonstrate a wide range in the sensitivity of the 𝛿 ¹⁸O_sw latitudinal gradient to different processes. In order of importance, the parameters that produce the most change in the high-to-low latitude surface gradient are decreased Arctic river runoff, closure of the Bering Strait, decreased overturning circulation, and decreased atmospheric water vapor transport. Based on these results, we conclude that using the latitudinal Atlantic 𝛿 ¹⁸O_sw gradient to constrain overturning rates would require better estimates of Arctic runoff change and smaller uncertainties for LGM 𝛿 ¹⁸O_sw measurements.

Waelbroeck et al. (2014), Constraints on surface seawater oxygen isotope change between the Last Glacial Maximum and the Late Holocene, Quaternary Science Reviews, 105 102-111.

Hourdin, F., et al. (2013) Climate Dynamics 40: 2167.

Notz, et al. (2013) Arctic sea-ice evolution as modeled by Max Planck Institute for Meteorology's Earth system model Journal of Advances in Modeling Earth Systems Volume 5, Issue 2 173-194.

Schmidt, G. A., et al. (2014), Configuration and assessment of the GISS Mode lE2 contributions to the CMIP5 archive, J. Adv. Model. Earth Syst., 6, 141-184.