A Dynamical Model Of ENSO Prediction Based On Tropical-Extratropical Precursor Signals

El Ni?o-Southern Oscillation(ENSO)dominates the time scale climate change from season to year in the tropical Pacific.Under the background of global warming,various extreme weather events are constantly occurring,which greatly affects the progress of national economy and social development.Therefore,the successful prediction of ENSO event is not only an international frontier issue in the field of atmospheric science,but also a major national demand for the government to actively respond to climate change.In the past few decades,researchers have made great progress in the theoretical prediction of ENSO.However,the existing prediction models of ENSO still have poor prediction skills when they begin to predict in spring or when they forecast across spring.This spring forecasting barrier for ENSO remains to be resolved.What is more noteworthy is that after 2000,the prediction skills of most ENSO prediction models have significantly decreased,and even there are phenomena of omission and misstatement.Therefore,improving the prediction model to eliminate or reduce the obstacle of spring prediction is the top priority in the current ENSO theoretical prediction research.In this paper,according to the scientific problems,starting from the physical dynamics that result in the change of ENSO mechanism,analyzes the possible factors influencing the spring forecast barriers,optimizing extra-tropical SST precursor factor(north Pacific Victoria modal(VM)and the south Pacific quadrapole(SPQ),the kinetics of ENSO prediction model is established,the main conclusions obtained are as follows:(1)For the spring predictability barrier phenomenon of ENSO,this paper firstly studied the influence of the VM on the persistent barrier of ENSO SST.It is found that the VM can prolong the persistence of ENSO SST,especially enhance the self-persistence of ENSO SST in spring,and reduce the intensity of persistence barrier phenomenon.The main reasons for these results are that the VM can affect the intensity and phase transition of westerly anomalies in the western equatorial Pacific(5°N–5°S,150°E–230°W),thus affecting the charging and discharging process of ENSO.When the strong VM and the strong ENSO events occurred simultaneously,the intensity of the anomalous westerliers in the equatorial western Pacific ware weaker than when the weak VM and the strong ENSO events occurred simultaneously.The weakened westerliers will further affect the charging and discharging mechanism of ENSO.In addition,we found that two consecutive El Ni?o events were also associated with strong VM events.(2)To solve the problem of how the extra-tropical SST signals affect the warm pool in the western Pacific,the changes in the warm pool area(WPA)range and the zonal position of the eastern boundary(WPEB)are determined by analyzing the dynamic process of VM affecting the warm pool in the western Pacific.Analysis results show that when the spring VM is positive(negative)phase,on the summer cause central equatorial Pacific SST anomaly warming,sea surface temperature(cold),at the same time the western equatorial Pacific will produce abnormal west(east)wind,thus strengthen(weaken)the eastern equatorial Kelvin waves,make the equator ocean surface warm water(cold water)to the west to the east,make WPEB moving east(moving westward),WPA expanded(narrow).Secondly,VM can further influence the expansion of WPA and the zonal migration of WPEB by affecting subsurface SST in the equatorial western Pacific.In addition,the results show that the spring VM index can effectively be used as a good predictor of WPA range size and WPEB zonal position about a year in advance.(3)For the phenomenon of the significant decrease of prediction skills of most ENSO prediction model after 2000,we analyszed the changes of relationship between ENSO and tropical-extratropical precursors during pre-2000 and post-2000.Futhermore,we also took the extratropical precursory signals into an ENSO prediction model,building a model which contains tropical and extratropical factor of ENSO.Notably that the prediction technique of our ENSO prediction model is the highest in 12-10 in advance,and the R between the prediction results and the observed Ni?o3.4 index are above 0.7,and the RMSE is less than 0.8.In addition,the phenomenon of spring predictability barrier of the original model can effectively reduced after took the extratropical signals into consideration.