Relationships between Wintertime Sea Ice Cover in the Barents Sea and Ocean Temperature Anomalies in the Era of Satellite Observations

Investigation of the predictability of sea ice cover in the Barents Sea is of paramount importance since sea ice changes in this part of the Arctic not only affect local marine ecosystems and human activities but may also influence weather and climate in northern midlatitudes. Here, observational data from the period 1981-2018 are used to identify statistical linkages of wintertime sea ice cover in the Barents Sea region to preceding sea surface temperature (SST) and Atlantic Water temperature anomalies in that region. We find that the ocean temperature anomalies formed by local air-sea interactions during the winter-to-spring season are a significant source of predictability for sea ice area (SIA) in the Barents Sea region the following winter. Optimal areas for constructing SST predictors of Barents Sea SIA and skill scores from retrospective statistical forecasts are shown to differ between the periods up to and since the onset of rapid sea ice decline in the region. In the EARLY period (1982-2003), springtime SSTs in the western Barents Sea predicted 44% of the variance of the following winter Barents Sea SIA. In the LATE period (2003-17), springtime SSTs in the southern Barents Sea predicted 70% of the variance of the following winter Barents Sea SIA. Regression analysis suggests that feedbacks from anomalous winds may be important for the predictability of wintertime sea ice cover in the Barents Sea region.