ALBERT

Description: The projected increases in Earth’s mean temperature entail substantial changes in the climate’s variability. Appropriate mitigation and adaptation measures require understanding of these changes since they impact for example the occurrence of extreme events. While projections of the global mean temperature are relatively well constrained based on the employed forcing scenario, they are highly uncertain with respect to the hydrological cycle and local temperature variability. Thus, examining how periods of past warming can help constrain these changes is of vital importance.To this end, we examine how the variability of surface temperature and precipitation changes in simulations of past and future warming from climate models of varying complexity and compare these with changes in proxy-based reconstructions. Based on the simulations, we analyze the moments of the distributions of temperature and precipitation (variance, skewness and kurtosis), as well as the power spectra with a focus on societally-relevant annual to centennial timescales. The analysis contrasts the projected changes under future warming scenarios with those found in transient simulations of the Last Deglaciation from models ranging from an energy balance model to Earth System Models. Changes observed in the simulations of the Last Deglaciation often highly depend on timescale and forcings, in particular changing volcanic activity, meltwater release and ice distributions alter patterns of variability. Based on this, we examine how the faster rate of future change impacts, and potentially limits, the conclusions to be drawn about future climatic changes based on past periods of warming.