Diverse and Weak Simulated Stratospheric Responses to Future Arctic Sea‐Ice Loss

Abstract Climate models project that the Arctic Ocean could see an ice‐free summer by the middle of this century. Through coordinated simulations, the Polar Amplification Model Intercomparison Project (PAMIP) aims to elucidate the causes and consequences of future Arctic sea‐ice loss. There is particular interest in understanding the mechanisms by which midlatitude weather and climate may be impacted, including via the “stratospheric pathway”; Arctic sea‐ice loss and associated warming is proposed to induce a wave‐driven weakening of the wintertime stratospheric polar vortex, which could subsequently impact tropospheric circulation. However, this is not well understood: studies do not so far find a robust stratospheric response to Arctic sea‐ice loss in either strength or sign. Here, we conduct novel analysis of the stratospheric response in thirteen PAMIP‐contributing models, looking beyond the typical time‐ and zonal‐mean diagnostics. Although our results overall confirm the lack of robust response, one model, HadGEM3‐GC31‐MM, has a statistically significant equatorward shift in vortex latitude, deceleration of vortex winds, and increase in sudden stratospheric warmings. Its response is found to be highly state‐dependent, significant only in the easterly phase of the quasi‐biennial oscillation (QBO). Though we cannot comprehensively conclude why models simulate this range of responses, our analysis does highlight areas for consideration in future work to better constrain the stratospheric response to Arctic sea‐ice loss. We explore the role of ensemble size, resolution and basic state, including zonal‐mean winds in the polar and midlatitude stratosphere and upper troposphere, as well as the QBO.