Diagnosing non-Gaussian temperature distribution tails using a back-trajectory analysis

Abstract:

Changes in extreme temperatures due to global warming will be manifested in more complex ways if the underlying temperature distribution is non-Gaussian compared to if it were Gaussian. Such non-normality in temperature distribution shape is common globally, with shorter-than-Gaussian tails leading to more rapid changes in the frequency of extremes under uniform warming than Gaussian or long-tailed distributions. Given the potential impact of non-Gaussian tails under future climate conditions, it is important to diagnose the underlying conditions and processes that lead to distribution asymmetry in the current climate. This work employs back-trajectories and associated large-scale circulation to gain insight into the underlying physical drivers of tail shape in a selection of mid-latitude locations with non-Gaussian temperature distribution tails. Although there is variability in characteristics of trajectories and the upstream environment at select locations, results reveal principal pathways for air parcel propagation associated with preferred patterns in synoptic-scale circulation. Locations exhibiting long cold tails require rare meteorological conditions to transport the coldest air from poleward source regions owing to local topography and prevailing winds, whereas extreme warm temperatures at short-tailed locations are regulated by marine air advection. Significant relationships with recurrent modes of atmospheric and sea surface temperature variability further suggest the influence of teleconnection wave patterns and ocean temperatures on extreme temperature occurrence over land, though non-advection-based processes are also important in some cases. Results provide added insight into our understanding of temperature extremes and how they may change in the future at regional scales.