Regional hotspots of temperature extremes under 1.5?°C and 2?°C of global mean warming

摘要

Local- and regional-scale heat extremes can increase at a significantly greater rate than global mean changes, presenting challenges for human health, infrastructure, industry and ecosystems. We examine changes in regional absolute temperature extremes for a suite of global regions under 1.5?°C and 2?°C of warming above pre-industrial levels, as described by the Paris Agreement. We focus on area-average values of observed monthly averages of daily maximum and minimum temperatures in 12 regions and calculate the most extreme monthly records observed. Next, using a large ensemble (HAPPI; Half a Degree Additional warming, Prognosis, and Projected Impacts) of decade-long simulations both of the present day and stabilised at these higher warming thresholds, we explore how changes in temperature extremes temperatures scale with global mean warming in these timeslice simulations. In the models, we focus on the 99th percentile values of monthly maximum temperatures and the 1st percentile of the monthly minimum temperatures. We define and identify hotspots of warming for various global mean warming levels, where projected changes in regional extremes are greater than global mean temperature changes. We identify overall hotspots of extremes, which are regions where the tail of the temperature distribution (above 99th percentile) warms at a faster rate than the rest of the temperature distribution in response to mean global warming increase. For monthly maximum temperatures, Central Europe, North Asia, West and East North America experience the greatest projected increases in extremes relative to means, and for monthly minimum temperatures, Central, West, East and North Asia, and East North America are identified as extremes hotspots. Although the scaling of increasing extremes with global mean temperatures is regionally variable, all regions benefit from the reduced severity of monthly maximum temperatures under lower global warming thresholds.