Toward a mechanistic characterisation of marine heatwaves

Mechanistic understanding of marine heatwaves (MHWs) requires a suitable definition for their detection, an approach to characterise their evolution, and an effective method to understand their causality. Much of our recent knowledge regarding MHWs has been achieved using a point-wise statistical definition that quantitatively defines MHWs as measurable warm ocean temperature extremes relative to a given threshold. While this commonly used definition is easy to use, with MHWs readily detectable and with near-global coverage from satellite sea surface temperature data, it does not quantify the spatial scale of events, their evolution in space and time, nor the association of that evolution with the key drivers. To overcome some of these limitations, more recent studies have investigated the evolution of MHWs as objects evolving in space and time to help broaden our understanding of MHWs. Our new approach represents an important step toward mechanistically characterising the space and time evolution of MHWs – it not only builds upon and extends object-based kinematic studies of MHWs but additionally connects these spatiotemporally evolving MHWs with their key drivers. Finally, we examine the potential predictability of these MHWs based on a linear inverse modelling approach.