Quantifying the state-dependent causal effect of Barents-Kara Sea ice loss on the stratospheric polar vortex in a large ensemble simulation.

Abstract

The Barents-Kara Sea ice concentration (BKS) has undergone dramatic declines in recent decades, consistent with the overall reduction in sea ice across the Arctic region. There has been a long-standing scientific question whether this BKS loss significantly influences winter temperature extremes over mid-to-high latitudes. While there is ongoing debate on this point, it is generally acknowledged that BKS loss affects the stratospheric polar vortex (SPV) through the enhancement of upward propagating waves, which itself can subsequently influence surface weather and climate conditions. However, due to the large internal variability within the climate system and the limited observational data, the strength of the BKS-SPV linkage and its dependence on different background states remain unclear. In this work, we investigate the causal effect of BKS change on SPV using a climate model with large ensemble simulations. Consistent with previous literature, the results indicate that BKS loss significantly weakens the SPV, with the magnitude of the response varying with El Niño-Southern Oscillation (ENSO) and Quasi-Biennial Oscillation (QBO) phases, indicating a state-dependent causal effect. In particular, El Niño is found to suppress the causal effect of BKS change on the SPV, whereas La Niña and neutral ENSO strengthen it, which is consistent with what is found from observations. In contrast, the effect of QBO alone is relatively weak but becomes more pronounced when combined with ENSO. Dynamical analyses reveal that both tropospheric wave forcing and modulation of stratospheric wave propagation contribute to the state-dependent causal effects. By leveraging large ensemble simulations and combining statistical and physical analyses, this study provides an additional perspective on understanding the factors influencing the SPV response to BKS loss, which could ultimately impact surface climate.