Most climatological studies characterize the future climate change as the evolution between a fixed current baseline and the future. However, as climate continues to change, ecosystems and societies will need to continuously adapt to a moving target. Here, we consider indicators of the pace of temperature change estimated from CMIP5 projections of an ensemble of climate models. We define the pace as a difference in relevant metrics between two successive 20-year periods, i.e. with a continually moving baseline. Under the strongest emission pathway (RCP8.5), the warming rate strongly increases, and peaks before 2080. All latitudes experience at least a doubling in the warming rate compared to the current period. Significant shifts in temperature distributions above twice the standard deviation between two successive 20-year periods expand from 9 % of continents on average currently to 41 % by 2060 onwards. In these regions, a warm year with a return period of about 50 years would become quite common 20 years later. The fraction of the world population exposed to such shifts will grow from 8 % to about 60 % on average, i.e. 6 billion people. Tropical areas are strongly affected, especially West Africa and South-East Asia. Low mitigation (RCP6.0) limits the warming rate to current values. Medium mitigation (RCP4.5) even reduces population exposure to significant shifts in temperature distributions to negligible values by the end of the century. Strong mitigation (RCP2.6) is the only option that generates a return to values similar to the historical period for all our indicators related to the pace of temperature change. This alternative way to analyze climate projections can yield new insights for the climate impacts and adaptation communities.
