Quantifying the physical processes leading to atmospheric hot extremes at a global scale

Atmospheric hot extremes are amongst the deadliest climate hazards and a focal point of the public and scientific discourse on climate change. Yet, the relative importance of the involved physical processes, temperature advection, adiabatic warming due to subsidence, and diabatic heating, is still debated. Here, we resolve this controversy by quantifying the contributions of these processes to near surface temperature anomalies during the hottest days of the years 1979–2020 in the ERA5 data set (hereafter TX1day events) at a global scale. To this end, a novel temperature anomaly decomposition is developed which evaluates the Lagrangian temperature anomaly equation (derived from the thermodynamic energy equation) along kinematic backward trajectories. We first use this decomposition to show that the extreme near surface temperature anomalies (hereafter T') during the June 2021 heat wave in western North America were primarily produced by diabatic heating, and, to a smaller extend, by adiabatic warming, while advection did not contribute significantly. Then, we systematically decompose T' during TX1day events globally and find that their composition strongly varies geographically. Advection dominates over midlatitude oceans, adiabatic warming near mountain ranges, and diabatic heating over tropical and subtropical land masses. However, in many regions TX1day anomalies arise from a combination of these processes. The time and spatial scales of their formation are 60 hours and 1100 km in the global mean, respectively, with large variability. The formation of these extremes is, therefore, inherently non-local and features distinct formation pathways in different regions, which implies hitherto poorly explored mechanisms for changing magnitudes of hot extremes in a warming climate.