Download files
Abstract
Oxygen isotope (δ18O) records from sediment cores, ice cores, and speleothems are a key proxy for reconstructing past climate. Since they are subject to changes in land ice, sea ice, ocean circulation, precipitation, evaporation, river discharge, and temperature, the interpretation of these records has been a major challenge. Here, an oxygen-isotope-enabled Earth System Model is used to partially resolve the complexity of δ18O records.
In Chapter 2, idealized experiments of a Heinrich Stadial are performed and compared with 36 marine sediment cores. Planktic δ18O is influenced by the volume and isotopic signature of the meltwater, changes in oceanic circulation and climate as well as calcification temperature, while benthic δ18O is mostly affected by local seawater temperature changes. Furthermore, advection of 18O-depleted surface waters during the recovery of the Atlantic Meridional Overturning Circulation (AMOC) leads to a δ18O decrease in the deep North Atlantic. This is in contrast to a hypothesis that links such decrease to brine rejection during sea ice formation.
Chapter 3 includes a detailed analysis of the lags in meltwater signal propagation, changes in ocean circulation, and the temperature effect in benthic records during a Heinrich Stadial. Time lags of several thousand years are found for meltwater signal propagation in the deep ocean. However, these lags have only a minor effect on δ18O anomalies during Heinrich Stadials. A comparison between two different modes of stadial ocean circulation shows large differences in deep ocean δ18O anomalies as recorded by calcifying foraminifera. These anomalies primarily reflect changes in deep ocean temperatures, which are caused by changes in global water masses.
Finally, in Chapter 4, transient simulations of Marine Isotope Stage 3 are conducted and compared with records from ice cores, ocean sediment cores, and a cave speleothem. The relatively good agreement between the simulated timeseries and paleoproxy records supports the hypothesis that stadial-interstadial cycles were driven by changes in AMOC. The simulated δ18O anomalies differ significantly between Heinrich stadials and non-Heinrich stadials, due to the volume of the 18O-depleted meltwater added and the corresponding changes in oceanic temperature, circulation and climate.