Atmospheric Waves and Regional Weather Extremes

点击翻译按钮获取中文摘要

Literature Review and Recent Progress Heatwaves, cold snaps, flooding and other extreme weather can have a devastating impact on humanity, with heightened suffering and mortality, as well as economic losses. For example, the 2010 Russian heatwave caused ~55,000 deaths, a 25% reduction in annual crop yield, and ~US$15 billion in economic losses. Clearly, skillful predictions of such events with lead-times of weeks to months can provide huge benefits to humankind; unfortunately, our present abilities to do this are extremely limited. Large-scale atmospheric waves are associated with many such events. Recently, I have shown that such waves are connected with particular patterns of atmospheric flow: atmospheric waveguides. These waveguides may provide a source of probabilistic seasonal to sub-seasonal (S2S) predictability. Objectives and HQP My overarching long-term research objective is to increase understanding and predictability, and reduce the impacts of extratropical extreme weather events, by determining how the probability of such events is influenced by large-scale atmospheric flow and associated compounding factors. Within the scope of this Discovery Grant research program progress towards this goal will be reached through a series of studies led by 9 MSc and PhD HQP under my mentorship. I will actively recruit HQP from under-represented and disadvantaged groups, and consider HQP applications from a holistic perspective. The work is organized into four short-term objectives: 1. Investigation of connections between extreme events and large-scale atmospheric flow and related predictability in present-day and future climates; 2. Quantifying effects of compounding factors on the probability and prediction of extreme events in present-day and future climates; 3. Quantifying and understanding prediction skill and biases in state-of-the-art S2S dynamical forecasts of extreme events; and 4. Developing and validating empirical and hybrid empirical-dynamical S2S probabilistic extreme events forecasts. Methodology This work will be accomplished using quantitative statistical and machine-learning analysis on gridded observational data, as well as ensembles of climate model and forecast simulations. Causal effect networks will be used to establish causal links, allowing a deeper investigation of the underlying dynamics of our climate system. Established causal links will then be used to develop empirical probabilistic S2S forecasts. Impact Significantly, this research program will forward our understanding of the atmospheric dynamics associated with extreme events, allowing a deeper understanding of model biases, and how to mitigate them. A hybrid empirical-dynamical early warning system will be developed for extreme events over Canada to provide more accurate probabilistic forecasts of extreme events at longer lead times than is currently available. Collaboration with end-users will help provide targeted and tailored information to maximise impact.