I’m pleased to announce that I’ve just started a new Post-Doc today at the University of Southampton, funded by the EPSRC/ReCoVER Network’s Early Career Researcher fund. I’ll be working for the next few months as a Research Fellow at Ocean and Earth Sciences in the National Oceanography Centre on a project entitled “Can early warning signals be reliably detected in the Cenozoic palaeoclimate record?” along with Co-Is Toby Tyrrell, James Dyke, and Tim Lenton. This will continue with research on the same topic that I started during my PhD, with the aim of extending the data and techniques used and write up some papers.
Some background: there are many points in Earth’s history where the Earth System is hypothesised to pass a ‘tipping point’ beyond which a rapid transition to a new and very different state occurs. These critical transitions are common in other complex dynamical systems and are often preceded in datasets by ‘early warning signals’ (EWS) such as critical slowing down (i.e. the system’s recovery time in response to perturbations slows down) and increasing variability (as the data gradually contains more extreme values). Dakos et al. [2008] and subsequent studies found that EWS can be detected prior to several past climate shifts, suggesting that critical transitions can successfully be detected in the palaeorecord and that palaeo tipping points can be identified. However, doubts have been raised about the reliability of EWS analysis on palaeoclimate records, the degree to which parameter selection can affect the results, and the risk of committing the ‘prosecutor’s fallacy’ when analysing suspected critical transitions. In my PhD I did a pilot study in which I analysed the highest-resolution palaeorecords currently available across a number of perturbations to the Cenozoic carbon-climate system, and found some promising results even when using a cautious approach to counter potential problems. In this Post-Doc I’ll focus on these most promising events with additional analytical techniques and data and publish the results in due course.
I’d like to thank the ReCoVER network and EPSRC for funding this research, and I look forward to sharing the results here in future!
Dr. David Armstrong MᶜKay 