Abstract:The spring predictability barrier (SPB) of El Ni?o-Southern Oscillation (ENSO) is a difficult problem in the ENSO prediction. To understand how the dynamic and thermal factors affect the variability of sea surface temperature over the tropical Pacific Ocean during the spring is very important to understand the changes of SST in key area and resolve the SPB problem. In this work, a set of monthly data including sea surface wind stresses, sensible heat flux, latent heat flux, longwave radiation, shortwave radiation, and ocean current fields which are coordinated with each other during 1986 -2017 are generated by numerical model simulation. Based upon these data, we analyzed and diagnosed the dynamic and thermal influences and their contributions to the variability of the sea surface temperature (hereafter, ). The main results show as follows: (1) Compared with other seasons, the sea surface temperature presents a unique asymmetric seasonal shift that is from warming to cooling in the Ni?o3.4 area during the spring. It is due to a similar shift in the wind stress, net energy fluxes, and ocean current which have a robust relationship with . Further analyses demonstrate that the thermal effect plays an important role in the variability of local . Differently, the meridional advection always shows a negative contribution to the seasonal variability of . Meanwhile, the zonal advection terms turn into a cooling effect from a warming effect to during the spring, and the vertical advection effect is quite weak. (2) The interannual correlation between the tendency of anomaly and the dynamic/thermal effect shows that the thermal heating is positively associated with the Ni?o3.4 anomaly in the spring, as well as the zonal advection. However, the correlation between the meridional advection and the anomaly changes from negative to positive during the spring. (3) The quantitative analysis of the dynamic and thermal variance contributions in the Ni?o3.4 region also suggests that, the contribution rate of thermal effect is more than 50%, and the corresponding correlation coefficient is over 0.7. The contribution of zonal and meridional advection is about 10- 20%, respectively, but they are opposite to each other. And the other items provide less contribution.