Fingerprinting Heatwaves and Cold Spells and Assessing Their Response to Climate Change Using Large Deviation Theory

Extreme events provide relevant insights into the dynamics of climate and their understanding is key for mitigating the impact of climate variability and climate change. By applying large deviation theory to a state-of-the-art Earth system model, we define the climatology of persistent heatwaves and cold spells in key target geographical regions by estimating the rate functions for the surface temperature, and we assess the impact of increasing ${\mathrm{CO}}_2$ concentration on such persistent anomalies. Hence, we can better quantify the increasing hazard due to heatwaves in a warmer climate. We show that two 2010 high impact events—summer Russian heatwave and winter Dzud in Mongolia—are associated with atmospheric patterns that are exceptional compared to the typical ones but typical compared to the climatology of extremes. Their dynamics is encoded in the natural variability of the climate. Finally, we propose and test an approximate formula for the return times of large and persistent temperature fluctuations from easily accessible statistical properties.

Figure 1

Summer RFs for the ST for increasing averaging windows (see legend) for (a) Northwest, (b) Southwest, and (c) Southeast America, (d) the Mediterranean, (e) North Europe, (f) Northwest and (g) Northeast Asia, (h) the North Atlantic, and (i) the North Pacific. The black (blue dashed) lines represents RFs obtained via seasonal averages for the preindustrial (quadruple ${\mathrm{CO}}_2$) run, with thick (thin) lines showing empirical estimates (quadratic fits). On top the mean ST and $\tau$ for the control (black) and quadruple ${\mathrm{CO}}_2$ (blue) runs. The 95% confidence intervals of the rate functions—shaded areas—have been computed by bootstrapping the data year by year.

Figure 2

Same as Fig. 1 but for winter RFs.

Figure 3

2010 RHW. (a) Composite of ST anomaly fields from the model run with ST anomalies $\ge 4.5\mathrm{K}$ lasting 30 days in the locale indicated by the green dot. (b) Mean August 2010 anomaly fields (NCEP/NCAR). The color map (isolines) indicates the ST (500 hPa GPH) anomalies. See in Fig. 11 of the Supplemental Material [35] the events contributing to the composite in (a).

Figure 4

2010 Russian heatwave. (a) Composite of precipitation anomaly fields constructed as in Fig. 3. (b) Mean August 2010 observed anomaly fields (CRU TS4 dataset).

Figure 5

Same as Fig. 3 but for the 2010 MD. See in Fig. 13 of the Supplemental Material [35] the events contributing to the composite in (a).