CFSv2-Based Statistical Prediction for Seasonal Accumulated Cyclone Energy (ACE) over the Western North Pacific

A hybrid dynamical-statistical model is developed for predicting the accumulated cyclone energy (ACE)a measure that can synthesize genesis number, mean life span, and intensity of tropical cyclones (TCs)in the typhoon season (June-October) over the western North Pacific (WNP) using data from both observations and seasonal forecasts of the National Centers for Environmental Prediction's (NCEP's) Climate Forecast System, version 2 (CFSv2). The model is built on the relationships between the observed ACE and the large-scale variables for the period of 1982-2010. Four predictors are selected based on both previous work in the literature and statistical analyses in this study, including vertical zonal wind shear over the equatorial western North Pacific (Ushear), sea surface temperature (SST) gradient (SSTG) between the southwestern Pacific (SWP) and the western Pacific warm pool, Nino-3.4 SST, and SWP SST. Based on the cross validation, the hybrid model is finally constructed with the combination of the summer Nino-3.4 and SWP SST at the 4-to-2-month lead (January-March) and the summer Ushear and the April SSTG at the 1-to-0-month lead (April-May). The hybrid model is shown to be skillful in predicting WNP seasonal ACE starting from January, with the correlation coefficient ranging between 0.58 and 0.81 and the root-mean-square error ranging between 1.26 and 0.91 (scaled by 10(5) m(2) s(-2)) initiated from January to May. The prediction experiments for 2011-13 using the hybrid dynamical-statistical model showed better skill and longer leads than that using the pure statistical models.