WWBs, ENSO predictability, the spring barrier and extreme events. Issue 17 (2nd September 2014)

• Record Type: Journal Article
• Authors: Lopez, Hosmay; Kirtman, Ben P.
• Is Part Of: Journal of geophysical research. Volume 119:Issue 17(2014)
• Page Start: 10, 114
• Page End: 10, 138
• Publication Date: 2014-09-02
• Subjects: Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1002/2014JD021908
• Languages: English
• ISSNs: 2169-897X

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>This study examines how semi‐stochastic Westerly Wind Bursts (WWBs) affect El Niño Southern Oscillation (ENSO) predictability. An ensemble ENSO prediction experiment is presented in which the Community Climate System Model version 3 (CCSM3) and CCSM3 with a state‐dependent WWB parameterization are used as both "truth" and as predictor systems. Inclusion of WWBs has little effect on ENSO predictability if the "truth" lacks WWBs. If the "truth" includes WWBs, the limit of ENSO predictability is larger for a forecast system that captures the correct <italic>statistics</italic> of WWBs. Predictability drops considerably if a forecast system that lacks WWB events is used to predict a "truth" that includes WWBs. At longer lead times, predictability is more dependent on the dynamical properties of the truth; that is, the importance of capturing the WWB statistics becomes less important and the statistics (e.g., signal‐to‐noise ratio) of the truth determine the limit of predictability. At short leads, ENSO predictability depends on the prediction system and the "truth." ENSO prediction skill is model and phase dependent. Predictability of extreme warm events remains a challenge as the number of ensemble members required to capture these events is on the order of 100 members. Finally, we examine real ENSO predictions with and without the WWB parameterization. It is found that including WWBs in the prediction system significantly increases ENSO prediction skill compared with a prediction system that lacks WWBs. Also, it is found that the so‐called forecast spring prediction barrier is, at least partially, caused by the lack of WWB representation in the forecast system.</p> </abstract>