Attribution of NAO Predictive Skill Beyond 2 Weeks in Boreal Winter

Weeks 3–6 averaged winter North Atlantic Oscillation (NAO) predictive skill in a state‐of‐the‐art coupled climate prediction system is attributed to two principle sources: upper and lower boundary conditions linked to the stratosphere and El Niño‐Southern Oscillation (ENSO), respectively. A 20‐membe...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Geophysical research letters 2020-11, Vol.47 (22), p.n/a
Hauptverfasser: Sun, Lantao, Perlwitz, Judith, Richter, Jadwiga H., Hoerling, Martin P., Hurrell, James W.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Weeks 3–6 averaged winter North Atlantic Oscillation (NAO) predictive skill in a state‐of‐the‐art coupled climate prediction system is attributed to two principle sources: upper and lower boundary conditions linked to the stratosphere and El Niño‐Southern Oscillation (ENSO), respectively. A 20‐member ensemble of 45‐day reforecasts over 1999–2015 is utilized, together with uninitialized simulations with the atmospheric component of the prediction system forced with observed radiative forcing and lower boundary conditions. NAO forecast skill for lead times out to 6 weeks is higher following extreme stratospheric polar vortex conditions (weak and strong vortex events) compared to neutral states. Enhanced weeks 3–6 NAO predictive skill for weak vortex events results primarily from stratospheric downward coupling to the troposphere, while enhanced skill for strong vortex events can be partly attributed to lower boundary forcing related to the ENSO phenomenon. Implications for forecast system development and improvement are discussed. Plain Language Summary Winter climate over Europe and eastern North America is significantly affected by variability of the North Atlantic Oscillation (NAO). In this study, we quantify the NAO predictive skill for lead times of 3–6 weeks and attribute it to two main sources: the stratosphere and the El Niño‐Southern Oscillation (ENSO) phenomenon. This is done by contrasting ensembles of 45‐day reforecasts over 1999–2015 with the corresponding uninitialized atmosphere model simulations forced with observed radiative forcing, sea‐surface temperature, and sea ice conditions. We find that the model is able to better predict the NAO up to 3 weeks following extreme weak or strong states of the stratospheric polar vortex compared to stratospheric neutral vortex states. Enhanced weeks 3–6 NAO predictive skill for weak vortex events results primarily from stratospheric coupling to the troposphere, while enhanced skill for strong vortex events can be attributed in part to lower boundary forcing related to ENSO. These results have implications for forecast model development and improvement. Key Points Weeks 3–6 NAO predictive skill is attributed to stratospheric polar vortex conditions and ocean lower boundary forcing Enhanced NAO skill following weak vortex events results primarily from stratospheric coupling to the troposphere Enhanced NAO skill following strong vortex events can be partly attributed to ENSO
ISSN:0094-8276
1944-8007
DOI:10.1029/2020GL090451