Superiority of Mega‐ENSO Index in the Seasonal Prediction of Tropical Cyclone Activity over the Western North Pacific
Yuan Sun,Zhong,Tim Li,Lan Yi,Jian Shi,Yixuan Shen,Kefeng Liu,Yijia Hu,Zanmin Xu
DOI: https://doi.org/10.1029/2019ea001009
IF: 3.68
2020-01-01
Earth and Space Science
Abstract:Abstract In this study, we compared the performance of two potential predicators, that is, El Niño–Southern Oscillation (ENSO) index (Niño‐3.4) and mega‐ENSO index, in the seasonal forecast of tropical cyclone (TC) activity and its spatial distribution over the western North Pacific (WNP) during the extended TC season, which is of public concern. Our results clearly show that, although both mega‐ENSO and Niño‐3.4 indices in the preceding May are important predictors for the seasonal predication, the relative‐sea surface temperature (SST)‐dependent mega‐ENSO exhibits a higher skill in the seasonal forecasting compared with the absolute‐SST‐dependent ENSO. Further results show that, despite of stronger destructiveness of TCs in high mega‐ENSO (El Niño) years than in low mega‐ENSO (La Niña) years, more attention should be paid to the TCs in low mega‐ENSO years, which are more likely to occur in coastal areas compared with the TCs in high mega‐ENSO years. Due to the responses of TC genesis, TC potential intensity, and large‐scale flow to the SST change in low mega‐ENSO years, the WNP TCs tend to originate in the northwestern quadrant and intensify at high latitudes and then turn northwestward over the TC prevailing region, which contributes to the northwestward migration of the WNP TC exposure in terms of track density and destructiveness density and thus imposes more risks in the coastal areas in low mega‐ENSO years. In addition, despite the significant predication skill in forecasting TC activity when using mega‐ENSO/Niño‐3.4 as a single predicator, it is still far to predict reliable WNP activity, especially its spatial distribution, without considering other predictors.