Cause-and-Effect Relationships Between Sea Surface Temperature Changes in Different Regions During the Past 4.5 Million Years

Understanding long-term cause-and-effect relationships between changes at different locations and temporal scales is crucial for unraveling the key mechanisms that drive climate dynamics and their impacts on ecosystems and society. Here, we focus on deconvolving different modes of Sea Surface Temperature (SST) variability and the coupling between different ocean regions at timescales of orbital (Milankovitch) variability. To do so, we apply Multivariate Empirical Mode Decomposition (MEMD) and Linear Response Theory (LRT). Through MEMD, we identify significant SST amplitude modulations at Milankovitch timescales, correlations between oceanic regions, and a both spatially and temporally variable long-term cooling trend through the last 4 million years. Through LRT we assess nonlinear causal inference at different timescales, highlighting significant SST coupling between extratropical regions and both the Indian-Pacific Warm Pool and the Eastern Tropical Pacific at multi-millennial scales. Our methodology offers first insight into causal connections at different timescales and locations, and into how the large-scale redistribution of oceanic heat across latitudes, especially between equatorial and polar regions, potentially influences ocean circulation and, consequently, the observed pattern of SST fluctuations.