Analysis of Amplified Northern Hemisphere Persistent Flow Regimes

Weather regimes have been widely studied to identify recurrent weather patterns in specific regions, or less frequently, across the entire Northern Hemisphere (NH). Previous case study analysis by the presenters has found that persistent ridges across the eastern North Pacific Ocean and western North America can be linked to amplified flow patterns across the entire NH. Amplified persistent flow regimes (PFRs) across the NH can last for 1­–5 weeks, with major impacts to precipitation and temperature anomalies on subseasonal-to-seasonal (S2S) timescales. However, additional research is still needed to document the impact of synoptic-scale weather patterns on the formation of these long-duration PFRs. A better understanding of how a broader range of dynamical features can impact amplified, persistent upper-level flow patterns will hopefully expand our ability to generate forecasts on the S2S time scales. The goal of this presentation is to more clearly identify these NH PFRs, and understand the dynamical and thermodynamical drivers that lead to NH PFRs.

PFRs will be identified as periods of widespread persistent anomalies as defined by Miller et al. (2020) using CFSR reanalysis data for DFJ from 1979–2022. All PFRs will be aggregated using k-means clustering of the 500-hPa geopotential height data at the start of each PFR. Time-lag composite maps of various atmospheric fields (e.g. 250-hPa wind speed and geopotential height) for each cluster will be created to identify how far in advance and in which geographic locations the NH flow pattern begins to amplify before forming a PFR. Dynamical features such as anomalous tropical convection, baroclinic cyclones, atmospheric rivers, jet exit and entrance regions, warm conveyer belts, and Rossby wave breaking will all be considered. Within the composites, assessment of statistical significance through bootstrap resampling will objectively show regions of anomalous atmospheric conditions prior to PFR formation. This statistical significance testing will allow for a comparison of which dynamical features listed above are common amongst the PFR clusters and which are unique to each cluster.