Will today’s proxies work tomorrow? Revisiting the stationarity assumption for severe convective environments

Environmental proxies are widely used to estimate the occurrence of severe convective storms - such as hailstorms and supercells - in both present and future climates, as these storms are typically not resolved in climate models. These environmental proxies rely on atmospheric parameters, including convective available potential energy (CAPE), convective inhibition (CIN), low-level humidity, and vertical wind shear. A core assumption behind their application is climate stationarity: that similar thresholds of environmental conditions will continue to produce similar storm initiation and behavior. However, this assumption remains largely untested. In this study, we explicitly assess the stationarity of convective proxies under climate change using convection-permitting, high-resolution climate simulations over Europe for the present-day and a +3°C pseudo global warming scenario. A storm-tracking algorithm is applied to track supercells and hail storms and their (pre-storm) environments. This approach enables a detailed comparison of the convective environments, their lower bounds, and their projected future changes. Our results show marked changes in both storm characteristics and their environments, such as increases in mean precipitation, hail diameter associated with hailstorms and supercells, and CAPE, CIN and vertical wind shear in the storm inflow regions. Nonetheless, lower bounds on the distributions do not show marked changes. We further examine how shifts in the distribution of convective environments influence simple threshold-based proxies for hail and supercell occurrence, highlighting the implications of environmental changes on convection proxies.