Twentieth Century Seawater δ18O Dynamics and Implications for Coral‐Based Climate Reconstruction. Issue 6 (20th June 2018)

• Record Type: Journal Article
• Authors: Stevenson, S.; Powell, B.; Cobb, K.; Nusbaumer, J.; Merrifield, M.; Noone, D.
• Is Part Of: Paleoceanography and paleoclimatology. Volume 33:Issue 6(2018)
• Page Start: 606
• Page End: 625
• Publication Date: 2018-06-20
• Subjects: coral paleoclimate -- El Niño/Southern Oscillation -- isotope‐enabled modeling -- ocean modeling -- climate reconstruction; Paleoceanography -- Periodicals; Paleoclimatology -- Periodicals; 551.46
• DOI: 10.1029/2017PA003304
• Languages: English
• ISSNs: 2572-4517

Abstract

Abstract: The oxygen isotopic composition of tropical coral skeletons ( δ 18 O) is a crucial source of information on past El Niño/Southern Oscillation behavior. Both temperature and the δ 18 O of the surrounding seawater ( δ 18 Osw ) affect coral δ 18 O; a linear proportionality with δ 18 Osw is often used to infer past salinity variations, but the degree to which dynamical influences on δ 18 Osw may affect that relationship is still unclear. Here we use the isotope‐enabled Regional Ocean Modeling System to investigate the dynamics of δ 18 Osw and salinity variations in different twentieth century climate regimes. The dominant modes of δ 18 Osw variability are the background trend and eastern/central Pacific El Niño, similar to salinity; likewise, budget analysis reveals a strong impact of ocean dynamics (both advection and vertical mixing/diffusion) on δ 18 Osw variations during El Niño and La Niña events. These dynamics lead to alterations in the δ 18 Osw :salinity relationship across the Pacific: the linear approximation is most accurate near the eastern edge of the western Pacific warm pool but with nonlinearities due to large δ 18 Osw excursions during El Niño. At other sites, the δ 18 Osw :salinity relationship has substantial scatter and explains relatively little δ 18 Osw variance. Strikingly, interannual variability can lead to site‐specific δ 18 Osw :salinity regression coefficient changes of up to 0.4‰/psu on multiyear time scales. This calls the reliability of coral‐based estimates of El Niño/Southern Oscillation extremes into question and highlights the need for increased dynamical understanding as well as sustained in situ observations of δ 18 Osw and salinity variability. Plain Language Summary: Tropical corals record past El Niño and La Niña events through changes in the isotopic ratios of oxygen in their skeletons, which is affected by the isotopic composition of seawater local to the reef. This seawater δ 18 O is thought to be closely related to salinity and is used to reconstruct past salinity variations associated with El Niño. However, this paper uses an isotope‐enabled ocean model to show that salinity and seawater δ 18 O have a complicated relationship. The processes affecting the two variables can be quite different in places, and their relationship varies significantly with time; therefore, it may be much more difficult to use coral records to reconstruct past salinity changes (and thus, El Niño impacts) than was previously thought. Key Points: Salinity and seawater δ 18 O variations are not always well correlated, and their relationship is affected by ENSO dynamics The salinity:seawater δ 18 O relationship varies strongly on interannual time scales Reconstructing salinity based on seawater δ 18 O may have large uncertainties