Identification Of The Internal Multidecadal Variability Of SST And Its Impact On Air Temperature And Rainfall Over Land

Based on the Pre-industrial control simulation,historical all-forcing simulation,historical greenhouse-gas-forcing simulation,historical natural-forcing simulation and historical miscellaneous-forcing simulation from seven coupled models from the Coupled Model Intercomparison Project Phase 5(CMIP5),several methods including linear trend(LT),global mean SST(GM),single model multi-members ensemble mean(SMM),multi-models multimembers ensemble mean(MMM),ensemble empirical mode decomposition(EEMD)and empirical orthogonal function(EOF),have been used to separate the internal multidecadal variability(IMV)of SST.It is found that the EEMD is the most effective method to identify the internal multidecadal variability of SST in the Atlantic(AIMV).The second are LT and SMM methods.The GM method works differently in different models,but it is generally not as good as LT and SMM methods.EOF method is the worst in separate the IMV in the models.EEMD method has been used to isolate the AIMV in the observations and to study its influence on the air temperature and precipitation over land.It has been found that the North Atlantic sub-polar region can trigger abnormal high pressure and influence the air temperature and precipitation in other regions.During the positive phase of the AIMV,the air temperature in Jun.-Jul.-Aug(JJA)decreases and the precipitation increases in the central North America and the northern and southwestern China,while the air temperature increases and precipitation decrease in the northern South American.In Dec.-Jan.-Feb(DJF),the air temperature increases in the western North America and eastern Australia,while it decreases in the western Australia and most of China.The precipitation in DJF increases in the western North America,western Australia and most of China,while it decreases in the eastern Australia.EEMD method has also been employed to isolate the IMV in the Pacific in the observations and to study its influence on the air temperature and precipitation over land.New indices are defined to represent the internal multidecadal variability in North Pacific(NPIMV)and South Pacific(SPIMV),respectively,which are independent of each other.When the NPIMV is in warm phase,the near-surface air temperature in DJF is colder than climatology in the north of North America and most of the east and central Siberia,but are warmer than climatology over Tibetan Plateau;the precipitation is more than average in the northwest and northeast of North America,but is less than average in Mexico and the region from northern India to southwest of China.Corresponding to a warm SPIMV,the near-surface air temperature in JJA is warmer in most of Europe,East Russia and north of North America,but is colder in west Siberia;the precipitation is more in central Russia but less in west Russia.It is through teleconnection wave train that NPIMV and SPIMV exert influence on the air temperature and precipitation in remote regions.