A Mechanism for the Arctic Sea Ice Spring Predictability Barrier.

The decline of Arctic sea ice extent has created a pressing need for accurate seasonal predictions of regional summer sea ice. Recent work has shown evidence for an Arctic sea ice spring predictability barrier, which may impose a sharp limit on regional forecasts initialized prior to spring. However, the physical mechanism for this barrier has remained elusive. In this work, we perform a daily sea ice mass (SIM) budget analysis in large ensemble experiments from two global climate models to investigate the mechanisms that underpin the spring predictability barrier. We find that predictability is limited in winter months by synoptically driven SIM export and negative feedbacks from sea ice growth. The spring barrier results from a sharp increase in predictability at melt onset, when ice‐albedo feedbacks act to enhance and persist the preexisting export‐generated mass anomaly. These results imply that ice thickness observations collected after melt onset are particularly critical for summer Arctic sea ice predictions. Plain Language Summary: Observations over the past 40 years have documented a significant decline in Arctic sea ice extent and thickness. These rapid changes and their implications for Northern communities, shipping industries, wildlife, fisheries, and natural resource industries have created an emerging operational need for regional summer sea ice predictions. This study is motivated by the following question: How far in advance can accurate predictions of regional summer sea ice be made? Recent work has shown evidence for an Arctic sea ice spring predictability barrier, which may fundamentally limit the accuracy of predictions made before May. However, the physical mechanism for this barrier has remained elusive. In this study, we investigate this mechanism using a sea ice mass (SIM) budget analysis, which allows for a process‐based attribution of summer sea ice predictability. We considerthe relative roles of ice growth and melt (thermodynamics) and ice motion (dynamics) in determining the spring predictability barrier. We find that predictability is limited by ice motion and growth in winter and increases rapidly in spring due to melt processes. These results imply that ice thickness observations collected after spring melt onset are particularly critical for summer Arctic sea ice predictions. Key Points: The Arctic sea ice spring predictability barrier is investigated using a daily mass budget analysisA mechanism for the spring predictability barrier is proposed, involving three distinct predictability regimesThe spring barrier is expected to shift earlier under Arctic warming due to shifts in melt onset timing [ABSTRACT FROM AUTHOR]