Скачать или смотреть CMOS Toronto Centre - Building Confidence in Regional Projections of Extreme Precipitation

Dr. Neil Tandon, York University

With rising temperatures, models and observations have shown increasing extreme precipitation intensity over most of the globe. This correspondence follows from basic thermodynamic principles: as atmospheric temperature increases, the equilibrium vapor pressure of water also increases in accordance with the Clausius-Clapeyron equation, which means that there is more atmospheric moisture available to form precipitation. Regionally, however, changes in dynamics—specifically vertical velocity of air—can act to greatly amplify or completely oppose the extreme precipitation increase expected from thermodynamic principles. Thus, building confidence in regional projections of extreme precipitation intensity requires understanding the reasons for changes in vertical velocity during extreme precipitation events, referred to as “extreme ascent.” In this talk, I will go over recent and current efforts to understand long-term changes in extreme ascent and challenges with accurately modelling extreme ascent. I will include analysis of global climate models (GCMs) as well as results from high-resolution regional simulations and discuss the trade-offs with each approach. Past studies based on analysis of GCM output have suggested a variety of reasons for changes in extreme ascent, and my group has been performing controlled experiments with regional models to try to clarify matters. These experiments suggest that a horizontal transport process called “differential vorticity advection” plays a key role in driving extreme ascent changes, but important questions remain about the influence of convective parameterization (model approximations of small-scale convection). Thus, future efforts to build confidence in projections of extreme precipitation will likely require improvements to convective parameterization. Alongside these efforts, increases in computational capability will allow greater usage of global convection-resolving models that do not require convective parameterization.