A New Perspective on Hydrographic and Sea Ice Variability in the Labrador Sea

As perennial sea ice vanishes from the Arctic Ocean, attention turns towards improving our understanding of seasonal sea ice variability. Seasonal sea variability in the Labrador Sea and Baffin Bay is not fully understood despite expressing 27% of Northen Hemisphere seasonal ice variability and having significant climatological importance (e.g., altering deep convection propensity, storm track location, CO2 sequestration rates, etc.). Leading seasonal ice variability hypotheses postulate large-scale atmospheric variability (e.g., NAO) as the primary driver - an interpretation mainly due to scarce high latitude in situ oceanographic data.

A novel perspective on the role of hydrographic variability in controlling seasonal sea ice maxima is revealed by one-year reconstructions of the three-dimensional time-varying sea ice-ocean state. These reconstructions are syntheses of in situ and satellite-based oceanographic and sea ice data with a state-of-the-art 30 km coupled ice-ocean model. Model and data are brought into consistency in a least-squares sense using the adjoint method (aka method of Lagrange Multipliers or 4DVAR) within the ECCO state estimation framework.

Analysis of these reconstructions reveals a first-order role of upper-ocean salinity anomalies for the development of ice across the thermohaline front separating cold, fresh Arctic Waters on the Baffin Bay and Labrador Shelves and the warm salty Irminger Waters in the central Labrador Sea. The enhancement of upper-ocean stratification from large positive buoyancy fluxes associated with ice meltwater release is found to be critical within the low-salinity anomaly region but ineffectual once the ice edge encounters convectively-entrained Irminger Waters ventilating in the mixed layer.