Analyses Of The Impacts Of Extratropical Atmosphere In The North Pacific On The SST Around Tropical Eastern Pacific During ENSO Transition Phase

Using the composite analysis,correlation analysis and SVD methods,the impacts of extratropical atmosphere in the North Pacific on the sea surface temperature(SST)around tropical eastern Pacific during ENSO transition phase are analyzed by using the global atmospheric reanalysis datasets from NCEP/NCAR,ERSST.V4 from NOAA,ORAS4 from ECMWF and datasets from 7 models in the Phase 5 of Coupled Model Intercomparison Project(CMIP5)historical simulation experiments.The present study consists of the following several parts:1?Based on reanalysis datasets,the first part analyzes the correlations between extratropical atmosphere in the North Pacific on the sea surface temperature(SST)around tropical eastern Pacific using sea level pressure(SLP)during ENSO transition phase in E-L type,in which La Ni?a occurs in the following year of El Ni?o,and E-N type,in which no La Ni?a occurs in the following year of El Ni?o.The results are as follows:(1)The positive SLPA of the key area(155°W-130°W,10°N-30°N)in June are strongly correlated with the negative SSTA around tropical eastern Pacific over the same phase and later.The probability of the positive SLPA in the key area for the E-L type is greater than the E-N type.(2)The results between the SLP in the North Pacific from March to June and the SST in December by means of SVD methods also confirm the negative correlation between the SLPA in the key area and the SSTA around tropical eastern Pacific.(3)The propagation of the positive phase of OKJ wave train in June may contribute to the positive SLPA in the key area.(4)The positive SLPA in the key area in June may result in more southward surface current velocity anomalies around the key area from July to September,which may contribute to transmit more cold water from the middle latitude to the equator.It may cause the negative SSTA around tropical eastern Pacific in December,to some extent,which may be conducive to the formation of La Ni?a.2?Based on 1956-2005 monthly reanalysis datasets from NCEP,the second part evaluates 7 models in CMIP5 historical simulation experiments to find the models which are better in simulating the statistical atmospheric and marine characteristics during ENSO transition phase.The results are as follows:(1)According to the EOF analyses,the first modes of 7 models are close to the observation which is related with ENSO,in terms of variance contribution,intensity and spatial variation despite of some differences.Their second modes represent the increasing trend of SST.(2)All of the 7 models can simulate the transition characteristics of the positive and negative phase of ENSO,but El Ni?o events in most models are relatively less frequent than the observation,while La Ni?a events in all of the models are more frequent than that.For the E-L type,the frequencies in four models are higher than the observation.However,for the E-N type,the frequencies in most models are lower than the observation.(3)In terms of the standard deviation,the maximum and minimum values of Ni?o3.4 index,NorESM1-M is better at reproducing SST,while the simulated SST in CCSM4,CESM1-CAM5 and CESM1-WACCM is higher than the observation,GISS-E2-H and GISS-E2-R lower.In the E-L type,El Ni?o events come to an end in March or April Y(0),and La Ni?a events begin in July or August Y(0).Therefore,May and June are the transition period,which is a little late for the observation.(4)From the SVD results,it finds that the SLPA in the significantly positive area of the subtropical North Pacific in June relate to the SSTA around tropical eastern Pacific in December.NorESM1-M is better at simulating this feature compared with other models.CCSM4?CESM1-CAM5?CESM1-WACCM and GISS-E2-H can only reproduce partial characteristics.As for May,the first mode of CESM1-WACCM represents that abnormal centers,which correspond with the positive phase of OKJ wave train,are relevant to the SSTA around tropical eastern Pacific in December,which is consistent to the observation,and the spread of the positive phase of the wave train is connected with the negative SSTA around tropical eastern Pacific,which may contribute to La Ni?a.(5)In sum,NorESM1-M is the best one at simulating the statistical atmospheric and marine characteristics during ENSO transition phase compared with others.