Accelerated western European heatwave trends linked to more-persistent double jets over Eurasia

Abstract

Persistent heat extremes can have severe impacts on ecosystems and societies, including excess mortality, wildfires, and harvest failures. Here we identify Europe as a heatwave hotspot, exhibiting upward trends that are three-to-four times faster compared to the rest of the northern midlatitudes over the past 42 years. This accelerated trend is linked to atmospheric dynamical changes via an increase in the frequency and persistence of double jet stream states over Eurasia. We find that double jet occurrences are particularly important for western European heatwaves, explaining up to 35% of temperature variability. The upward trend in the persistence of double jet events explains almost all of the accelerated heatwave trend in western Europe, and about 30% of it over the extended European region. Those findings provide evidence that in addition to thermodynamical drivers, atmospheric dynamical changes have contributed to the increased rate of European heatwaves, with implications for risk management and potential adaptation strategies.

Fig. 1. Increasing heatwave trends over the midlatitudes and Europe.

a Decadal trends in heatwave frequency (days/decade) and b heatwave cumulative intensity (°C/decade) for July-August 1979–2020. c Probability density distributions of decadal trends of heatwave frequency of all land grid points for Europe (in dark red, as the region included in the dashed box of (a, b): 35–70°N and 10oW-50°E) and the midlatitudes (20–70°N) excluding Europe (in blue) and d probability density distributions of decadal trends of heatwave cumulative intensity. The mean trend for each distribution is shown with dashed vertical lines and provided on the top right of the panels. The continuous vertical lines correspond to 0 (i.e. no trend). The two distributions were compared for each case with a Kolmogorov-Smirnov test (p values shown on the center-right).

Fig. 2. Jet stream states and surface temperature.

a–c Clusters of the vertical profile of the zonal (averaged over the Eurasian domain) mean zonal wind (u, shading) with frequency of occurrence provided in parenthesis. The climatological mean of the zonal mean zonal wind for the whole period is plotted with dashed contours (plotted from 5 to 20 m/s every 5 m/s). d–f Frequency (gray line) and maximum persistence (orange line) of each cluster per year. The decadal linear trend and respective p-values are given for both time series on the top left of the panels. g–i Anomaly composites of (linearly detrended) zonal wind at the 250hPa pressure level (u250) for each cluster (shading). The climatological mean of the zonal wind at 250hPa is plotted with dashed contours (plotted from 5 to 25 m/s every 5 m/s). j–l Anomaly composites of (linearly detrended) mean surface temperature for each cluster. Anomalies in both cases are calculated with respect to daily climatology (to remove the seasonal cycle). m–o Composites of heatwave cumulative intensity (calculated after having removed the Tmax mean midlatitude-land trend from each grid point) shown as relative anomaly (%) compared to the climatology. All figures refer to the months of July-August of the period 1979–2020.

Fig. 3. Persistent double jets states.

a Composite of meridional winds at 250hPa (v250) for double jet events exceeding the 90th percentile of double jet persistence (i.e, events lasting more than 11 consecutive days, see Table S1 for the 20 most persistent events and their duration; shading). Contour lines show the v250 July-August climatology for the whole period 1979–2020 (plotted from −5 to 8 m/s every 2 m/s). b, c SOM cluster composites of v250 for persistent double jet events (frequency % of each SOM is shown in parenthesis). d, e Time series of daily similarity index for the two SOMs of v250 winds. The red dashed line shows a linear regression fit, with its slope and p-value plotted on the top left. The continuous red line shows a smoothed LOESS curve fit (span of 0.75). f, g Anomaly composites of (linearly detrended) mean surface temperature for each of the SOM clusters.

Fig. 4. Summers 1994, 2003, 2006, 2018: double jets and heatwave intensity.

a Anomalies of the 250hPa zonal wind (u250; shading, anomalies from climatology; dashed contour lines show the total wind speed climatology plotted every 5 m/s) for the longest double jet event in 1994: 23.07-19.08.1994. b Hovmöller diagram of the Eurasian (region seen in panel a) zonal mean zonal wind anomalies for July-August 1994 (shading; 5day running means centered on each day from 01.07-31.08.1994). The vertical dashed line refers to the first day of August and the red horizontal lines on the time axis show the days identified by the SOMs as double jets (dates for 1994: 1-12.07, 23.07-19.08). c Spatial distribution of heatwave cumulative intensity for July-August 1994. d As in a but for 21.07-18.08.2003. e As in b but for July-August 2003 (dates of all double jets for 2003: 11.07, 16-19.07, 21.07-18.08, 24-31.08). f As in c but for July-August 2003. g As in (a) but for 11-30.07.2006. h As in b but for July-August 2006 (dates of all double jets for 2006: 11-30.07, 12-16.08, 31.08). i As in c but for July-August 2006. j As in (a) but for 04-25.07.2018. k As in (b) but for July-August 2018 (dates of all double jets for 2018: 04-25.07, 19-20.08, 25-31.08). l As in c but for July-August 2018.

Fig. 5. Explained variance of heatwave cumulative intensity by double jet persistence.

a Explained variance (R²) per grid point of heatwave cumulative intensity based on linear regression on double jet persistence. Statistically significant coefficients (p < 0.05) are marked with black dots. b Scatter plots of heatwave cumulative intensity anomalies aggregated over all land grid points of the extended European domain (as seen in panel a and in the red dashed box of Fig. 1a, b) and double jet persistence. A linear fit (in red) and its confidence interval (dashed lines for the 5th and 95th percentiles obtained from 1000 bootstraps), R² (with 5th and 95th interval percentiles obtained from 1000 bootstraps in brackets), and p-values are shown on the top left of each plot. c As for b but with heatwave cumulative intensity aggregated only over land grid points with statistically significant coefficients in western Europe (dotted points in panel a within the region included in the dashed red box: 37–55˚N and 9˚W–14˚E). The linear trend of the time series was removed before the regression was applied in all cases.

Fig. 6. Observed versus estimated heatwave cumulative intensity trends based on the increase of double jet persistence.

a Observed residual trend (observed trend—mean midlatitude-land trend) in heatwave cumulative intensity (°C/decade) over Europe. b Estimated residual trend in heatwave cumulative intensity (°C/decade) over Europe. Grid points for which the estimated residual trend is of the same sign with the observed residual trend are marked with dots. Mean residual trend (°C/decade) for Europe and western Europe (region included in the dashed red box) is given on the top left of each panel.