Weakening and shift of the Arctic stratospheric polar vortex: Internal variability or forced response?

Recent studies have proposed that the Arctic stratospheric polar vortex has weakened and shifted away from the North Pole during the past three decades. Some of these studies suggest that this trend has been driven by a decline in Arctic sea ice leading to enhanced zonal wave number 1 waves propagating into the stratosphere and that it has in turn contributed to a recent wintertime surface cooling over North America and some parts of Eurasia. Here trends in several measures of the location and strength of the stratospheric polar vortex from 1980 to 2016 are examined in two reanalysis products. All measures show weakening and equatorward shift trends, but only one measure, the vortex centroid latitude, has a trend which is statistically significant at the 95% level in both reanalyses. By comparing large ensembles of historical simulations with preindustrial control simulations for two coupled climate models, the ensemble mean response of the vortex is found to be small relative to internal variability. There is also no relationship between sea ice decline and trends in either vortex location or strength. Despite this, individual ensemble members are found to have vortex trends similar to those observed, indicating that these trends may be primarily a result of natural internally generated climate variability. Plain Language Summary The Arctic stratospheric polar vortex is a region of strong winds between about 10 and 50 km in altitude encircling the North Pole during winter. Changes in the strength and location of the vortex can affect the weather below. Several recent studies have proposed that the vortex has shifted away from the pole and weakened from 1980 to 2016 as a result of declining sea ice in the Arctic. This study first looks at different measures for vortex changes using two observationally based data sets. Only one measure, the centroid latitude, has a trend which is statistically significant. Next it is investigated whether these trends can be attributed to anthropogenic changes (such as sea ice decline) or natural climate variability by using a large number of climate model simulations. Some of these simulations have increasing greenhouse gases, and others are fixed at preindustrial conditions. Despite large differences in sea ice between the two sets of simulations, there is not a large difference in the likelihood of vortex trends. This suggests that the weakening and shift of the vortex is primarily a result of natural internal climate variability.