Impact of Arctic Sea Ice Representation on Extended Medium-Range Forecasting: a Case Study of the 2016 Barents-Kara Sea Warming Event

Abstract

In early January 2016, Storm Frank, an extreme winter storm with a peak intensity of 928 hPa, intruded into the Atlantic sector of the Arctic. This led to unprecedented warming and significant sea ice loss in the Barents-Kara (B-K) Sea. Following this extreme warming event, a series of extreme weather events occurred in mid- and late-January across Eurasia, including a persistent blocking pattern near the Ural mountains and extreme cold wave events over Mongolia, China, and Korea. This study utilizes the Korean Integrated Model (KIM), coupled with an ocean-sea ice model, to reproduce this event and to examine its extended medium-range forecasting performance. While the control model effectively captures the initial Arctic warming, it struggles to reproduce the observed sustained warming that lasted over two weeks. Here, we identified that the model significantly overestimates the sea ice concentration in the B-K Sea, where the initial warming is more pronounced in observations. Through sensitivity experiments, we found that reducing the sea ice strength parameter, which governs the ice resistance to pressure and deformation, effectively alleviated this overestimation. This adjustment facilitates easier sea ice melting, strengthens the ocean-atmosphere interactions, and extends the duration of simulated Arctic warming. Our findings emphasize the crucial role of accurate Arctic sea ice representation in extended medium-range forecasting for East Asia, particularly for extreme weather events.