| Summary: | Amidst the ongoing climate crisis, concerns arise regarding the response of various components of the climate system, especially those vulnerable to abrupt shifts once a tipping point is reached. Predicting such behaviour is particularly challenging due to the unnoticeable changes before the transition. A promising alternative to assess the system stability and potentially warn of an incoming tipping point is based on detecting generic symptoms observed as a system gradually approaches the transition. The system experiences a slowdown in recovery from perturbations, or loss of resilience, leading to increased similarity and variability over time. This approach requires long-term, regularly spaced time-series, characteristics that are rare among observational records, especially in the ocean. The recent development of annually-resolved proxy records based on information encoded in bivalve shells offers a unique opportunity for assessing resilience in the marine environment. This thesis explores the potential of bivalve-derived reconstructions to assess changes in stability in the northern North Atlantic through two resilience indicators, lag-1 autocorrelation and variance. These explorations demonstrate the reliability of bivalve records to measure changes in resilience, particularly in autocorrelation, and provide guidelines for this purpose. An exploration of changes in resilience over the last millennium demonstrates that bivalves can effectively encode changes in stability. The analysis reveals that the subpolar gyre circulation system crossed a tipping point into the transition into the Little Ice Age and provides hints on how the input of freshwater from melting glaciers and sea-ice may have contributed to the destabilisation. Shifting the focus to recent times, an assessment of changes in resilience on a compilation of bivalve records across the northern North Atlantic indicates that the regional marine environment has lost resilience over recent decades. This destabilisation is likely linked to the ...
|
|---|