Based on the data form the monthly Sea Surface Temperature (SST) data of HadISST from 1951 to 2016, NCEP/NCAR reanalysis and Woods Hole Oceanographic Institution(WHOI) form 1958-2016, and statistical methods such as Empirical Orthogonal Function (EOF) and partial correlation analysis, the main characteristics of the North Atlantic Sea Surface Temperature Anomaly (SSTA) in spring and their association with the spring NAO and previously winter ENSO are studied. The results show that the first mode of EOF in spring exhibits a meridional tri-polar pattern, which revealed 35.7% variance. The spring tri-polar pattern SSTA over the North Atlantic is mainly affected by the spring NAO and the central and eastern tropical Pacific SSTA in the previous winter. After eliminating the influence of the previous winter Niño3.4, the partial correlation coefficient between the North Atlantic Tri-pole(NAT) mode index and the North Atlantic Oscillation (NAO) index is 0.50, which passes the 99% confidence level significance test. After eliminating the effects of spring NAO, the partial correlation coefficient between spring NAT index and the previous winter Niño3.4 index is -0.26, which passes the 95% confidence level significance test. Associated with the sea surface wind and sea surface turbulent heat flux anomalies in North Atlantic, the positive(negative) NAO in spring leads to the positive (negative) NAT SSTA. The previous winter ENSO events can cause atmospheric circulation anomaly and extratropical SST anomaly in spring, which in turn modulate the influence of spring NAO on the North Atlantic tri-polar pattern SSTA in spring.