Contrasting Atmospheric Drivers of Heatwaves and Heat Stress Extremes

Extreme heat stress events have a significant impact on society. They are typically identified as heatwaves, which are anomalously warm periods reportedly associated with large-scale persistent anticyclonic patterns. These complex and potentially compound events can be studied with multivariate impact-based indices. This is routinely done with known heatwaves and gives a new perspective focused on the social and ecological impact.

Nonetheless, it is unclear to what extent extreme heat stress spells coincide with extreme temperature events and what their dynamical and thermodynamical drivers are. There is still no comprehensive assessment of non-matching events, investigating their frequency, duration, and both large-scale and local processes involved.

In this work, we study heat stress events and heatwaves, with a focus on the non-overlapping cases and their leading meteorological conditions. To identify such episodes, we use global reanalysis data to compare some of the most common process-based definitions of heatwave with a 95th percentile threshold of Universal Thermal Climate Index (UTCI). Over a 40-year period, we estimate that more than 20% of high-UTCI days during the summer season do not correspond with equally extreme two-metre temperature (T2) conditions across much of Europe. The associated large-scale and mesoscale features are examined. We also analyze the daily cycle of heat stress, highlighting the presence of a phase difference between UTCI and T2 peaks. Furthermore, we investigate the use of nighttime UTCI in place of common T2 comfort thresholds to estimate nighttime thermal stress. Then, we compare case studies to assess the role of soil-temperature coupling and advection in determining heat stress extremes.

This work clarifies the difference between process-based and impact-based definitions concerning extreme heat spells and explores whether it is appropriate to pair them to capture events that might be otherwise neglected or underestimated.