Rewrite Super-Clausius–Clapeyron scaling of extreme precipitation explained by shift from stratiform to convective rain type as a headline for a science magazine post, using no more than 8 words

Global warming has been suggested to bring widespread precipitation intensification⁶, in particular regarding convective rainfall extremes⁵. Super-CC increases have served as a possible explanation for such anticipated intensification^4,32. The current work calls into question if the suggested super-CC increase in convective precipitation extremes actually has a mechanistic base by which thunderstorm cells come with invigorated intensities beyond the thermodynamic rate at higher temperatures^11,22,23. The present data unambiguously show that there is no exceedence of the CC rate at the scale of individual convective cells. To the contrary, the CC rate is a robust predictor of the change in convective precipitation intensity with temperature. Super-CC changes with temperature are found only as a statistical superposition of distinct rainfall types¹⁹.

Under such statistical superposition, precipitation clusters comprising both convective and stratiform precipitation, such as MCSs^24,26, could exceed CC scaling as a statistical ensemble. Societal impacts resulting from MCSs, such as flash flooding^2,3 in increasingly urbanized watersheds¹, could be amplified under a statistical super-CC scaling. Our analysis based on IDF curves indeed suggests that it is at the flash flood scale, on the order of 15–60 minutes, where storms show strongest super-CC increases. Probing whether convective fraction increases at similar exponential rates in state-of-the-art km-scale climate change simulations would be a logical next step. Such cloud-resolving simulations are also appropriate in detecting thunderstorm events in the spatio-temporal simulation output fields and conditionally analysing the scaling of convective vs stratiform contributions. Precipitation extremes may increase at different rates at different times of the year, as the partitioning into stratiform and convective contributions differs from season to season (Extended Data Fig. 10). Yet, current climate model projections may not accurately reproduce seasonal changes, as showcased by projections that underestimate the inland advection of wintertime maritime systems and the convective fraction of winter storms³³.

Given the current findings, refocusing the target of extreme precipitation modelling might be useful: with convective and stratiform components individually scaling along the thermodynamic CC rate, the focus could be shifted to unveiling the spatial organization of thunderstorms within mesoscale cloud fields. A strongly clustered thunderstorm population within a given MCS could locally lead to a severe flash flood whereas a scattered population would give rise to moderate, more widespread precipitation at the mesoscale. With the high-resolution simulation data now becoming available^34,35, detecting changes in clustering with temperature will be feasible. Prominent mechanisms for thunderstorm self-organization, such as cold pool interactions, are now heavily studied, often in idealized settings^{36,37,38,39,40,41}. Conceptual understanding, gained from such works, should find its way into realistic regional-scale studies to help inform future changes in organized convection, for example, convective processes within winter storms and large slow-moving summer systems, which may be more frequent in a warmer climate^33,42,43.