Interdecadal Transition In The Relationships Between The Western Pacific Subtropical High And Sea Surface Temperature And The Forecast Capacity Of Models

The western Pacific subtropical high(WPSH) and the sea surface temperatures all experienced marked interdecadal changes. However, whether the relationships between the WPSH and sea surface temperature forcings(SSTFs) also experienced an interdecadal transition remains unclear. Based on NCEP/NCAR reanalysis data, this study analyzes the abrupt interdecadal shift in the relationship between the WPSH and SSTFs. This study adopted a index to estimate the intensity of the WPSH by the first two empirical orthogonal function(EOF) modes of the 850-hPa geopotential height field. The results indicate that the interdecadal transition in the relationship the leading EOF modes and the SSTFs are obvious in approximately 1976-1977 and 1993-1994. Dividing the study period into three parts in terms of the year of the abrupt interdecadal shift, a comparative study of the anomalous circulation corresponding to the PC during the three periods is performed. The results reveal that the characteristics of SSTA and 850-hPa wind anomaly fields corresponding to the PCs are different from each other and exert an remarkably interdecadal change. The positive PC1 is closely related to the equatorial Eastern and Central Pacific(EECP) warming, the tropical Indian Ocean(TIO) warming and the Western North Pacific(WNP) cooling in spring and summer of three periods. The discharging process of “Indian Ocean Capacitor Effect” of tropical Indian Ocean in summer is suppressed by the downdraft in 1978-1992 and 1995-2013, causing the TIO warming has little influence on the WNPAC in summer. The positive PC2 correponds to significantly positive SSTAs in the TIO, the Maritime continent and the WNP, significantly negative SSTAs in the North Pacific in spring and summer and the EECP cooling in summer of three periods, while it is only closely related to the positive-phase NAT in spring of 1954-1975. An anticyclone over the North Pacific Ocean accelerates the phase shift of ENSO from El Ni?o to La Ni?a in spring of 1995-2013.The enhancement of WNPAC by the NAT is partly offset by the tropical North Atlantic warming with an positive tropospheric temperature anomaly in spring of 1954-1975.For the changes of East Asian summer circulation has great impact on the summer rainfall in China and the East Asian summer circulation is greatly influenced by tropical SSTA. Thus, it is necessary to understand the forecasting capacity of seasonal forecast models and the capability of describing relationship between the East Asian summer circulation and tropical SSTA more deeply. In this study, we evaluate the capacity of seasonal forecast models to forecast the intensity of East Asian Summer Monsoon(EASM) and WPSH in summer, basing on the reanalysis data from NCEP/NCAR and the model result of three seasonal forecast models from 1991-2013, namely CFS V2, BCC_CSM V2 and MRI-CGCM, which are from National Centers for Environmental Prediction(NCEP), National Climate Center(NCC) and Tokyo Climate Center(TCC) respectively. To illustrate the origin of model error, we analyze response capability of EASM and WPSH to tropical sea surface temperature(SST) anomaly in models, and the impacts of ENSO events on forecasting EASM and WPSH. Analysis results indicate that the forecast skill of EASM and WPSH are high in all models, while that of TCC model is relatively lower. An anomalous cyclone occurs over Western North Pacific, resulting in the intensity of EASM is stronger, while WPSH is weaker, than that of observation in all model. Meanwhile, the annual variability of them is less than that of observation. The characters of response of EASM and WPSH to tropical SST anomaly with seasonal evolution are close to that of observation in all models, while the response of EASM to the preceding tropical Pacific Ocean SST anomaly, as well as the preceding and simultaneous tropical Indian Ocean SST anomaly in NCEP model and TCC model is stronger than that of observation, and the response of EASM to the preceding and simultaneous tropical Pacific Ocean SST anomaly in NCC model is stronger than that of observation obviously. Besides, the response of WPSH to the preceding and simultaneous tropical Pacific Ocean, tropical Indian Ocean and tropical Atlantic Ocean SST anomaly in all models is stronger than that of observation obviously. The mean absolute error(MAE) of EASM and WPSH forecasted by three models in ENSO events is much less than that in normal years overall. The MAE of EASM and WPSH forecasted by NCEP model and NCC model are more in La Ni?a events than in El Ni?o events, while that forecasted by TCC is contrary, which indicates that ENSO events is an important source of forecast for summer circulation in East Asia.