Enhanced Northern Hemisphere Correlation Skill of Subseasonal Predictions in the Strong Negative Phase of the Arctic Oscillation

Abstract

AbstractThe Arctic Oscillation (AO) is the most dominant atmospheric variability in the Northern Hemisphere in boreal winter. Its negative phases sometimes bring extreme cold conditions over Eurasia and North America in boreal winter, impinging on various socioeconomic sectors. Thus, accurate prediction of the AO‐related conditions with a long lead time is greatly anticipated. This study investigates conditional prediction skill in the northern extratropics relative to AO phases using retrospective forecast data of multiple models provided by the Subseasonal to Seasonal Prediction Project, which is jointly conducted by the World Weather Research Programme and the World Climate Research Programme. We found that predictions starting from the strong negative AO phase tend to have enhanced prediction skill in terms of the anomaly correlation coefficient of 500‐hPa geopotential height, which measures the similarity of spatial patterns. The skill enhancement is not apparent in terms of the root mean square error score. We also discuss dynamical mechanisms behind the enhanced prediction skill. The Eliassen‐Palm flux diagnosis indicated that the strong negative AO phase induces the stronger eddy‐zonal flow feedback to sustain the anomalous zonal flow condition than the strong positive AO phase. Moreover, the anomalous zonal flow is associated with an anomalous zonally asymmetric pattern. As a result, the anomalous AO pattern is better predicted in the strong negative AO phase, contributing to the enhancement of the Northern Hemisphere correlation skill. Results highlight that dynamics inherent in the extratropical atmosphere can provide the subseasonal predictability in a certain atmospheric condition.