Haurwitz Memorial Lecture: Scale interactions and the generation of low-frequency variability in the atmosphere

Much of our understanding of low-frequency variability in the atmosphere is based on separation of the flow into zonal mean and eddies, or time mean and transient components, and subsequent analysis of the interactions between them. From this we find that much of the structure of atmosphere flow that appears in time mean maps and dominates variability on periods of weeks to months is driven by eddy, mean-flow interaction. Strong feedbacks between the mean flow and high-frequency transients give rise to both low-frequency variability and sensitivity of climatological features to small forcings such as incipient climate change. While these interactions are not linear, much of this can be understood on an intuitive level by applying linear theory. Mid-latitude jets are eddy-driven and their low-frequency variability is therefore enhanced by the interactions between variations of the low-frequency flow and the low-frequency envelope of forcing by high-frequency eddies. Low-frequency variability is greatest where the eddy-driven jets are strongest, such as in the North Atlantic and in the Southern Indian Ocean. The variability in eddy driven jets is asymmetric, with equatorward migrations of midlatitude jets being more persistent that poleward migrations of jets. The location of midlatitude jets in climate models is variable and models with more equatorward climatological jets also experience more poleward jet migration associated with global warming. The sphericity of the Earth seems to play a key role in these asymmetries by causing the turning latitude of poleward propagating wave to occur before waves can reach their critical latitude and break. This dramatically alters the momentum fluxes and zonal accelerations produced by the eddies. Zonal asymmetries and eddy feedbacks also greatly affect the influence of the QBO and ENSO on the stratospheric vortex in winter.