Changes in regional heatwave characteristics as a function of increasing global temperature

Abstract

The Paris Agreement calls for global warming to be limited to 1.5-2 °C. For the first time, this study investigates how different regional heatwave characteristics (intensity, frequency and duration) are projected to change relative to increasing global warming thresholds. Increases in heatwave days between 4-34 extra days per season are projected per °C of global warming. Some tropical regions could experience up to 120 extra heatwave days/season if 5 °C is reached. Increases in heatwave intensity are generally 0.5-1.5 °C above a given global warming threshold, however are higher over the Mediterranean and Central Asian regions. Between warming thresholds of 1.5 °C and 2.5 °C, the return intervals of intense heatwaves reduce by 2-3 fold. Heatwave duration is projected to increase by 2-10 days/°C, with larger changes over lower latitudes. Analysis of two climate model ensembles indicate that variation in the rate of heatwave changes is dependent on physical differences between different climate models, however internal climate variability bears considerable influence on the expected range of regional heatwave changes per warming threshold. The results of this study reiterate the potential for disastrous consequences associated with regional heatwaves if global mean warming is not limited to 2 degrees.

Figure 1

Median regression coefficients estimated from the CMIP5 model ensemble between global warming (°C) and seasonal (a) heatwave days; (b) number of events; (c) event duration; and (d) peak heatwave intensity. Created using NCAR Command Language (version 6.4.0) [Software]. (2017). http://dx.doi.org/10.5605/D6WD3XH5.

Figure 2

Regional (see Giorgi and Francisco, 2000) changes in heatwaves relative to 0.5 °C global warming thresholds estimated from the median of the CMIP5 model ensemble for (a) heatwave days; (b) number of events; (c) event duration; (d) peak intensity; and (e) regional mean warming. See table S.2 for region boundaries.

Figure 3

Spread in the regression coefficient estimated from CMIP5 models (left column) and the CESM ensemble (right column) for (a,b) heatwave days; (c,d) number of events; (e,f) event duration; and (g,h) peak intensity. The spread is given by the ensemble 1^st percentile subtracted from the ensemble 99^th percentile. Created using NCAR Command Language (version 6.4.0) [Software]. (2017). http://dx.doi.org/10.5605/D6WD3XH5.

Figure 4

Ensemble spread (99^th percentile – 1^st percentile) of increases in the number of heatwave projected by the CMIP5 ensemble (blue) and CESM ensemble (purple) per 0.5 °C global warming, for (a) Alaska; (b) the Mediterranean; (c) Australia; (d) East Asia; (e) the Amazon, and (f) East Africa. These regions were chosen as they are representative of all 21 regions analysed. Note that the spread due to internal variability (estimated from CESM) is reasonably consistent across the thresholds. See table S.2 for region boundaries.