Global Low-frequency Teleconnection Decadal Variation Studies

Atmospheric teleconnection, with good persistency and stability, is the principal model in terms of global circulation at low frequency. The paramount observation have already demonstrated that the occurrences of the low-frequency wave at specific geographic areas will lead to global circulation anomalies, thus resulting in climate anomalies and disasters at corresponding locations, or even at global scale. Therefore, the studies of the temporal-spatial variation of the teleconnection have long been a hot topic because of the great theoretical implication and practical application (e.g. disaster prevention and reduction)In this study, interdecadal variability of basic mode of global zonal-mean atmospheric circulation is investigated by using multi-year observations with coherent wavelet analysis. The results showed that interdecadal variability with 16a period clearly exists in the basic mode of global zonal-mean atmospheric circulation. Also, the change sequence has been revealed by the coherence phase. On the time scale above 16a, the change of zonal-mean temperature leads the zonal-mean flow and the zonal-mean geopotential height about 3 and 1 years, respectively; Also, the change of zonal-mean geopotential height leads the zonal-mean flow about 2 years. The increase (decrease) of zonal-mean temperature will cause the zonal-mean geopotential height descending (ascending) at high-latitude area and ascending (descending) at mid-latitude area; then further causes the zonal-mean westerly accelerating (decelerating) at mid- and high-latitude and tropics, and westerly decelerating (accelerating) at polar regions and subtropics.The temporal-spatial characteristics of 30-60d low frequency variation of global 500-hPa geopotential height field has been studied with various statistics methods (e.g. Buttworth filter, EOF, etc.). The results showed that the dominant mode of the low-frequency variation at 500hPa are the Pacific/North American (PNA),East Atlantic (EA),Polar/Eurasia, and the Antarctic Oscillation (AAO). Here after, these four global low frequency modes are referred as to GLT. Additionally, in order to describe the temporal variability the GLT Index (GLTI) is defined. It is found that GLTI has obvious interdecadal characteristics, i.e., GLTI remains low (high) value before (after) mid 1970s, corresponding to GLT's negative (positive) phase. The relationship between the GLT and 500-hPa zonal-mean temperature, zonal-mean flow, sea surface temperature and sea ice has been studied. The results showed that the external forcing zonal-mean flow and GLT are closely related on the scale of above the 18a, and their changes have a sequence on such scale. The SST and Sea ice change first, then the zonal-mean flow changes, and finally the GLT changes. Combined with the global 500-hPa zonal-mean flow distribution, the conclusion can be drawn for the interdecadal variation of external forcing zonal-mean flow and GLT:PDO pattern transforming from a negative phase to a positive phase /sea ice cover decreasing will cause the zonal-mean westerly accelerating at mid- and high-latitude, and westerly decelerating at polar reigns and subtropics. The corresponding GLT pattern transforms from a negative to a positive phase. Therefore, SST and sea ice, as the main external forcing, exert influences to GLT through changing the 500-hPa zonal mean flow distribution.The spatial spectrum coefficients variability, which reflects the basic climate mode difference before/after the GLT interdecadal variation, is analyzed by utilizing the spatial spectrum method. The results showed that the basic climate mode, especially the zonal-mean flow, changes dramatically, as the response to the interdecadal variation of the external forcing. Also, the influence of the basic climate mode to the instability growth mode of low frequency variation is studied via barotropic primitive equation model. Furthermore, the mechanism of interdecadal variation of GLT is obtained. The results illustrated that both quasi 15d and 60d instability mode can be excited under different basic flows, but with quite different e-folding time. The e-folding time of quasi 15d and 60d decreases dramatically after the interdecadal variation. In addition, the corresponding instability increasing rate increases too. This finding is consistent with the observations.The impact of SST and sea ice on GLT has been further approved by the numerical simulation. The GLTI behaves quite different under positive/negative PDO SST pattern..Specifically, GLTI is almost positive/negative under the forcing of the positive/negative PDO phase SST, which is consistent with the observations. The distribution of 500hPa geopotential height over the mid-latitude and high-latitude in the Northern Hemisphere is approximately opposite. Also, the zonal symmetric structure is obvious. It is found that the primary characteristics of 500hPa geopotential height in the Southern Hemisphere are similar to those in the Northern Hemisphere. Compared to the observations, the characteristics of symmetric distribution of the positive/negative centers of the zonal mean-low are more noticeable between the positive/negative PDO SST forcing. However, it is worth mentioning that the spatial distribution of the zonal mean flow in the Southern Hemisphere does not agree with the observations well. In comparison, the simulation in the Northern Hemisphere is consistent with the observations, except that the location of the center is toward equator. The results indicate that PDO pattern SST anomaly is possibly responsible for the interdecadal variation of GLT in the Northern Hemisphere, but possibly not the reason of the interdecadal variation in the Southern Hemisphere. The small/large GLTI (i.e., negative/positive GLI) corresponds with negative/positive SAC anomaly, which is consistent with observations. The 500hPa zonal-mean flow behaves quite different under the the positive/negative SAC. That is, the spatial distribution of the zonal mean flow changes from "+,-,+,-,+,-,+"to"-,+,-,+,-,+,-" along with SAC transforming from positive to negative, or vice versa In a nutshell, the sea ice is one of the most prominent factors causing the interdecadal variation of GLT.Thus, we argue that the interdecadal variation of SST and Sea ice cause the interdecadal variation of general circulation through changing the basic climate state, especially changing the zonal-mean flow, and thus leads to the differences of instability modes, i.e., weak (strong) instability mode before (after) interdecadal variation. Additionally, GLTI remains low (high) before (after) interdecadal variation, corresponding with the negative (positive) GLT spatial distribution.