Constraining Amazonian land surface temperature sensitivity to precipitation and the probability of forest dieback

Abstract The complete or partial collapse of the forests Amazonia is consistently named as one top ten possible tipping points Planet Earth in a changing climate. However, apart from few observational studies that showed increased mortality after severe droughts 2005 and 2010, evidence for such depends primarily on modelling. Such are notoriously deficient at predicting rainfall Amazon basin how vegetation interacts with poorly represented. Here, we use long-term surface-based observations air temperature to provide constraint modelled sensitivity changes precipitation. This emergent also allows us significantly constrain likelihood forest dieback. We conclude dieback under IPCC scenario RCP8.5 (crossing point) not likely occur twenty-first century.