UC San Diego

As an emerging proxy for total ocean heat content, atmospheric noble gas ratios in ice cores have the promise to provide key insights into Earth’s energy budget, and how it has responded to past climate perturbations. This method takes advantage of the mass-conservation of krypton, xenon, and dinitrogen in the combined ocean and atmosphere, and their unique temperature- dependent solubilities in seawater. By measuring changes in the noble gas ratios of the well-mixed atmosphere, we can deduce their change in the ocean due to warming or cooling, which allows us to track changes in total ocean heat content. As a new area of research, there is still much to learn about the proxy’s potential, as well as its limitations. The motivation of this dissertation is twofold: to identify the main sources of uncertainty in mean ocean temperature reconstructions and to apply the mean ocean temperature method to previously unstudied climate intervals.

To address our first goal, we focus on processes that decouple the noble gas ratios measured in ice cores from mean ocean temperature. While past studies have focused on this decoupling at the ocean-atmosphere interface (e.g. through dissolved gas disequilibrium), we primarily focus on the decoupling of gases at the atmosphere-ice sheet interface, and within the ice sheet itself. We identify previously unappreciated or underappreciated mechanisms of gas fractionation that, if not appropriately accounted for, introduce error into mean ocean temperature records. Based on our findings, we outline specific steps to identify and correct for these processes in order to minimize error in future mean ocean temperature studies.

To address our second goal, we reconstruct mean ocean temperature with ice cores from the Taylor Glacier blue ice area. Taylor Glacier has the advantage of being recently-drilled and clathrate-free, making it an ideal site for mean ocean temperature reconstruction. Using well- established methods of blue ice dating based on gas synchronization with well-dated ice cores, we identify and collect ice cores from the last interglacial period and last glacial inception for mean ocean temperature reconstruction. Considering the mean ocean temperature records presented here, along with previous reconstructions, a pattern emerges from the data; over a range of timescales and background climate states, we find a dominant role of the ocean’s overturning circulation in controlling the timing and magnitude of mean ocean temperature change.