Freeze-up and Break-up of Arctic Sea Ice in the U.S. Beaufort and Chukchi Seas (2019-2020)

After freeze-up investigations from 2009-10 through 2016-17, a multi-year gap preceded this study, making it essential to re-establish continuity in the freeze-up dataset and confirm whether the warming, storm-frequency, and multi-year ice trends identified through 2016-17 had persisted or shifted in the intervening years. This ninth study in the series, commissioned by BSEE, was designed to document the 2019-20 freeze-up season in the Alaskan Beaufort and Chukchi Seas by describing evolving ice conditions, mapping features and pile-up events, correlating ice canopy changes with meteorological conditions, and using the combined 2009-17 and 2019-20 dataset to characterize present-day freeze-up processes relative to the 1980s. Researchers combined open-source and proprietary meteorological data, ice charts, drift buoy data, and RADARSAT-2 satellite imagery collected from September 2019 through February 2020 with four aerial reconnaissance flights conducted in late February 2020. The 2019-20 winter was warm by historical standards but unexceptional within the recent record, producing first-year ice thickness (162 cm in the Beaufort, 148 cm in the Chukchi) that ranked in the middle of the 11-winter dataset; storm activity in both seas reached historical minima for the study series; and multi-year ice invaded the nearshore Beaufort for the second consecutive year while also entering the Chukchi via the established multi-year gateway. The extended dataset confirms accelerating trends: freeze-up now arrives 2.3 days per year later in the Beaufort and 4.6 days per year later in the Chukchi, implying a substantially longer open-water season ahead; first-year ice thickness has declined 15% since the early 1980s; and while flaw lead frequency remains stable, multi-year ice invasion probability is now quantified at about 20% per season in the Beaufort and 65% in the Chukchi based on 20 years of data, sharpening the risk estimates used for offshore facility planning. This abstract was drafted with the assistance of Claude, July 2026.

Following the inaugural 2017 break-up study, continued investigation was needed to build a comparative dataset testing whether break-up in the Alaskan Beaufort and Chukchi Seas was truly trending earlier alongside the well-established trend of later freeze-up, and to refine methods—such as timing flights to each basin's break-up progress—based on lessons from the first study. This second break-up study, commissioned by BSEE, was designed to describe the 2020 break-up season, map ice features and pile-up/ride-up events, correlate ice canopy changes with meteorological conditions, and compare the 2017 and 2020 seasons with historical break-up patterns. Researchers combined open-source meteorological data, ice charts, drift buoy data, and RADARSAT-2 satellite imagery from May through July 2020 with two staggered sets of aerial reconnaissance flights—Chukchi flights in mid-June and Beaufort flights in early July—timed to each basin's break-up progress. The 2020 season was comparatively cool by recent standards in the Chukchi (ranking 35th of the past 51 years) despite an accelerated pace of landfast ice deterioration; large sections of ice up to 74 km long broke free from the Chukchi coast in a single event; and multi-year ice remained present throughout the study period in both seas at concentrations up to 70% in the pack ice, dispersing gradually as landfast ice disintegrated. Comparison of the 2017 and 2020 datasets against historical records supports the hypothesis that both break-up and the onset of open water are trending earlier, with open water advancing faster than break-up—shrinking the break-up season while lengthening the overall open-water season when combined with delayed freeze-up; storm frequency during break-up has nearly doubled since the early 1980s; and while multi-year ice residency during break-up appears to have declined since the 1980s, its continued presence confirms it remains a relevant hazard, with authors cautioning that more data are needed to confirm these early trends. This abstract was drafted with the assistance of Claude, July 2026.