Arctic Wintertime Sea Ice Breakup Events - Detection and Evolution

Abstract

Openings in the Arctic sea ice give rise to strong localized heat fluxes from the ocean to the atmosphere, causing ocean mixing and potentially leading to new sea ice growth. Despite the increased attention on sea ice breakup events due to an observed increase in their occurrence during winter, we still lack an automated sea ice breakup detection method. This thesis introduces the first automatic detection of wintertime pan-Arctic sea ice breakups from 2000 to 2018 using the spatiotemporal statistical analysis method Empirical Orthogonal Functions (EOFs) of extreme deformation anomalies obtained from the high-resolution sea ice model neXtSIM. The extreme anomalies are masked and filtered to find the dominant deformation variability explained by the three first EOF modes. Breakups in individual Arctic regions are defined from these modes and their corresponding 90th (P90) and 10th (P10) percentiles. A breakup event is defined when the amplitude variance of the temporal EOF exceeds P90 or falls below P10. Using ERA5 reanalysis, we present how synoptic weather patterns are congruent with extensive sea ice deformation and drift for the defined breakups in the individual regions. The atmospheric and sea ice evolution is studied for the days before, during, and after the onset day. Agreeing with previous findings, we find that the persistence and orientation of the wind relative to the coast are the key factors for sea ice breakup events in all Arctic regions. Alongshore flow is favorable for coastal lead formation, while offshore flow also favors coastal polynya formation.