In recent years, there has been more concern about cloud feedbacks. High-resolution climate simulations show that low-lying stratocumulus clouds will break up in a warmer climate, reducing their shading effect and allowing for greater warming.1 This is significant, because these clouds are common in the tropics, shading 20 per cent of low-latitude oceans.
Better modelling of such feedback mechanisms, especially through refinement against observational data, can help us understand these risks and improve the fidelity of our climate models. In recent years, for example, researchers have successfully reconstructed the history of the Atlantic Meridional Overturning Circulation (AMOC) going back over a century.2 This means it is now possible to put observations of current AMOC changes into their long-term context — and the data suggest that the AMOC is indeed slowing.3 This may, in the near term, increase the overall warming at the surface.4
Models also need to take more account of ecosystem feedbacks, such as those from the melting of permafrost (an event that could release large quantities of greenhouse gases, potentially accelerating and increasing the warming trend5), climate-induced human migration and coral bleaching, all of which can have feedback effects on climate systems.