Title: Contributions of the North Pacific Meridional Mode to Ensemble Spread of ENSO Prediction

Journal: Journal of Climate, 30:9167-9181, doi:10.1175/JCLI-D-17-0182.1

Authors: MA J.*, S. -P. Xie, and H. -M. Xu

Abstract: Seasonal prediction of El Niño–Southern Oscillation (ENSO) employs the ensemble method, which samples the uncertainty in initial conditions. While much attention has been given to the ensemble mean, the ensemble spread limits the reliability of the forecast. Spatiotemporal coevolution of intermember anomalies of sea surface temperature (SST) and low-level winds over the Pacific is examined in ensemble hindcasts. Two types of evolution of intermember SST anomalies in the equatorial Pacific are identified. The first features an apparent southwestward propagation of the SST spread from the subtropical northeastern Pacific southeast of Hawaii to the central equatorial Pacific in boreal winter–spring, indicative of the precursor effect of the North Pacific meridional mode (NPMM) on ENSO variability. Extratropical atmospheric variability generates ensemble spread in ENSO through wind–evaporation–SST (WES) in the subtropical northeastern Pacific and then Bjerknes feedback on the equator. In the second type, ensemble spread grows in the equatorial Pacific with a weak contribution from the subtropical southeastern Pacific in summer. Thus, the extratropical influence on ENSO evolution is much stronger in the Northern Hemisphere than in the Southern Hemisphere. The growth of Niño-4 SST ensemble spread shows a strong seasonality. In hindcasts initialized in September–March, the Niño-4 SST spread grows rapidly in January–April, stabilizes in May–June, and grows again in July–September. The rapid growth of the Niño-4 SST spread in January–April is due to the arrival of NPMM, while the slowdown in May–June and rapid growth in July–September are attributable primarily to the seasonality of equatorial ocean–atmosphere interaction. NPMM contributes to the ensemble spread in equatorial Pacific SST, limiting the reliability of ENSO prediction.