Reconstruction of historic sea ice conditions in a sub-Arctic lagoon

• Sea ice conditions were reconstruction based on local temperature record.
• A statistical model of ice conditions is shown to be useful to determine seasonally aggregate ice conditions.
• The methods are transferable to other locations.

Historical sea ice conditions were reconstructed for Izembek Lagoon, Bering Sea, Alaska. This lagoon is a crucial staging area during migration for numerous species of avian migrants and a major eelgrass (Zostera marina) area important to a variety of marine and terrestrial organisms, especially Pacific Flyway black brant geese (Branta bernicla nigricans). Ice cover is a common feature of the lagoon in winter, but appears to be declining, which has implications for eelgrass distribution and abundance, and its use by wildlife. We evaluated ice conditions from a model based on degree days, calibrated to satellite observations, to estimate distribution and long-term trends in ice conditions in Izembek Lagoon. Model results compared favorably with ground observations and 26 years of satellite data, allowing ice conditions to be reconstructed back to 1943. Specifically, periods of significant (limited access to eelgrass areas) and severe (almost complete ice coverage of the lagoon) ice conditions could be identified. The number of days of severe ice within a single season ranged from 0 (e.g., 2001) to ≥ 67 (e.g., 2000). We detected a slight long-term negative trend in ice conditions, superimposed on high inter-annual variability in seasonal aggregate ice conditions. Based on reconstructed ice conditions, the seasonally cumulative number of significant or severe ice days correlated linearly with mean air temperature from January until March. Further, air temperature at Izembek Lagoon was correlated with wind direction, suggesting that ice conditions in Izembek Lagoon were associated with synoptic-scale weather patterns. Methods employed in this analysis may be transferable to other coastal locations in the Arctic.