Atmospheric Dynamics and Impacts on Abrupt Arctic Summer Sea Ice Loss Events

Arctic sea ice extent (SIE) exhibits considerable year-to-year variability, with changes potentially amplifying as ice thins. Extremes of summer Arctic sea ice loss are intimately connected to the atmospheric forcing.  On sub-seasonal to seasonal time scales, there are significant correlations between upper-level heights, sea level pressure, and sea ice loss. On synoptic (i.e., ``weather'') time scales, individual synoptic cyclones are observed to be associated with rapid sea ice loss events. However, prediction of rapid sea ice loss is generally not very skillful beyond several days. Tropopause polar vortices (TPVs) are common, can be long-lived, and are necessary precursors for the formation of the surface cyclones responsible for sea ice loss.  This presentation explores the hypothesis that TPVs have significant impacts on the evolution and predictability of summer sea ice loss.  Thus, knowledge of TPV processes could have potential to improve predictions of sea ice loss.

This study employs an observational-based and a numerical modeling-based approach to test the hypothesis.  To establish the general characteristics of synoptic weather events and rapid sea ice loss over the Arctic, extreme summer sea ice loss events are identified from two-day changes in SIE 1979-2014 from the National Snow and Ice Data Center (NSIDC).  Data are filtered to include only contributions from individual and significant weather events using spectral analysis techniques, and extreme events are classified as the top 1\% of filtered two-day decreases in SIE. Atmospheric composites from ERA-interim over each of the extreme sea ice loss cases identified from 1979-2014 reveal the characteristics and variability of the synoptic-scale cyclones for these events, as well as the upper-level characteristics of TPVs present before the surface cyclones form.  Numerical modeling experiments focus on medium- to long-range predictions of the Arctic environment during the summer of 2006 and 2007, which are two years characterized by strongly contrasting cyclonic and anticyclonic pressure and tropospheric circulation anomalies and SIE.  To isolate the associated coupled, multi-scale interactions and atmospheric dynamics, experiments are designed using the non-hydrostatic dynamical core from the Model for Prediction Across Scales (MPAS). Results from an ensemble of atmosphere-only experiments indicate that forecast spread develops from the interaction of TPVs with amplifying Rossby waves.  Although the Rossby waves initiate in midlatitudes, high-resolution is most beneficial over the Arctic in order to better resolve the intensity of TPVs before wave interactions or surface cyclogenesis occurs.  Idealized sensitivity experiments confirm sensitivities of subsequent large-scale flow to TPV structure by artificially damping the local mechanisms that intensify TPVs.  These findings motivate consideration of atmopsheric, cryospheric, and oceanic couplings using MPAS coupled with the Community Earth System (CESM) model.