The Predictability of Northern Hemisphere Stratospheric Final Warmings and Their Surface Impacts

The Northern Hemisphere (NH) stratospheric polar vortex comes to an end each spring when the climatological westerly winds turn easterly, either because the stratospheric temperature gradient radiatively relaxes as sunlight returns to the pole or because the vortex has been dynamically disrupted and never recovers. Here we consider the predictability of NH final warmings in multi-model hindcasts from the Sub-seasonal to Seasonal (S2S) database. We compare the predictability of early warmings (which are largely dynamically driven) and late warmings (which are largely radiatively driven) and find that late warmings are predictable at longer lead times. We assess the spread in the ensemble members using a variety of different metrics to better understand the reason for the extended predictability of late final warmings. We also examine whether better forecasts of the final warming lead to improved skill of surface climate in the days following the observed event.