Vertical Velocity Shape and the Scaling of Tropical Precipitation Extremes with Warming

It is generally accepted that warmer climates lead to stronger tropical precipitation extremes. How different processes contribute to the scaling is still an area of active research, though, and the rate of increase of precipitation extremes with warming is poorly constrained. Several past studies have focused on contributions from a thermodynamic component – related to the temperature dependence of the saturation specific humidity lapse rate – and a dynamic component – related to changes in convective vertical velocities with warming. The combination of these two pieces has often been found to approximate the scaling of surface saturation specific humidity with surface temperature – or “Clausius-Clapeyron scaling”. However, it is unclear how robust this result is and how it might be affected by large-scale convective organization. Observations suggest a link between convective organization and the shape of updraft velocity profiles, and we use this connection to motivate developing theory for precipitation extremes in the case of updraft velocity profiles that are comprised of a combination of first and second baroclinic modes. Next, we use this theory to analyze updraft velocity profiles and precipitation scaling in two sets of cloud-resolving model (CRM) simulations. In one set, run in a narrow but 12,000 km long channel, convection self-aggregates over a wide range of surface temperatures and the simulations develop large-scale overturning circulations that resemble those found in the real tropics. In the other, run in a small square domain, convection does not aggregate. Projecting simulated updraft velocity profiles onto first and second baroclinic modes across a range of extreme precipitation quantiles provides us with a simple way to quantify changes in updraft velocity profiles with warming, and provides a framework for studying relationships between convection organization, updraft velocity profiles, and precipitation extremes in the CRM simulations.