The Strong Stratospheric Polar Vortex in March 2020 in Sub-Seasonal to Seasonal Models: Implications for Empirical Prediction of the Low Arctic Total Ozone Extreme

摘要

Using reanalysis and real-time forecasts from 11 sub-seasonal to seasonal (S2S) models, the strong stratospheric polar vortex in March 2020 coupled with Arctic ozone loss and their predictability are reported in this study. Arctic ozone reached a minimum value in March 2020 similar to the events in 1997 and 2011. Upward propagation of waves into the stratosphere in winter-early spring was anomalously weak, and the weakened Brewer-Dobson circulation transported less ozone-rich air to the Arctic stratosphere in November 2019-March 2020, which contributes around 40% of the total Arctic ozone loss in March, with the residual mainly explained by the chemical loss. Anomalous upwelling in the Arctic further cooled and intensified the polar vortex for ozone depletion and weak mixing, preventing recovery of ozone in early spring 2020. The polar vortex intensity and Arctic total ozone are strongly coupled, so a prediction of Arctic total ozone becomes possible for S2S models using metrics of the polar vortex as a proxy. The prediction of the strong stratospheric polar vortex in late March 2020 is assessed for four common initializations (February 27, March 5, March 12, and March 19) for S2S models. Displacement of the polar vortex toward the Western Hemisphere can only be forecasted in the two later initializations, and its intensity anomaly is underestimated in nearly all of those forecasts. Therefore, the empirical model using the S2S outputs also underestimates the Arctic total ozone. Despite this underestimation, this extremely strong polar vortex event led to improved surface predictability on subseasonal timescales: near-surface temperature and precipitation are well forecasted 2-3 weeks in advance.