Recognition Of Two Dominant Modes Of EASM Based On Multiple-monsoon Indexes And CMIP5 Model Results

Based on the ERA-Interim,NCAR/NCEP and JRA-55 reanalysis datasets?1979-2016?,the classification of 25 commonly used indexes of East Asian Summer Monsoon?EASM?has been investigated.On this basis,the relationship between two typical monsoon indexes and the dominant modes of EASM has been analyzed,and then the circulation pattern,climate anomalies and possible thermal driving factors corresponding to the dominant modes of EASM have been investigated.Finally,using the historical simulation data of 31 global climate models provided by the fifth phase of the Coupled Model Intercomparison Project?CMIP5?,the spatial structures,temporal variabilities and thermal driving factors of the dominant modes of EASM during 1979-2005 have been evaluated.The main conclusions are as follows:?1?The selected 25 monsoon indexes can be classified into two typical categories:I_EASMI1ASMI1 and I_EASMI2,which are independent of the dataset sources.The physical nature of two categories of monsoon indexes is associated with the dominant modes of EASM.I_EASMI1represents interannual mode of EASM,with a typical oscillation of 2-3 yeas,while I_EASMI2represents decadal mode of EASM,revealing a dominant period of 12 years.?2?I_EASMI1 is closely linked to the low-latitude circulation system and the anomalous circulation pattern is a typical EAP teleconnection pattern,which corresponds a typical tripole pattern of summer rainfall anomalies.The main driving factors of I_EASMI1 are the tropical sea surface temperature anomalies?SSTAs?,especially ENSO related SSTAs in precedent winter,and the summer SSTAs in the tropical Indian Ocean and the tropical Atlantic Ocean also play important roles in the abnormal activity of EASM.?3?I_EASMI2 mainly reflects the impacts of the middle-high latitude circulation system on the summer monsoon and is closely linked to a typical EU teleconnection pattern,which corresponds a dipole of summer rainfall anomalies.Further analysis suggests that the signal of anomalous monsoon activity mainly originates from the polar region and the middle and high latitudes of the Eurasian continent,which is closely related to the land surface thermal conditions of the Eurasian continent.?4?The models considered are able to simulate the spatial structures of both interannual and decadal mode of EASM,especially the interannual mode.However,the models generally lack the ability to simulate the temporal variabilities.Considering of spatial and temporal structures,INMCM4,GFDL-ESM2M,BNU-ESM,MPI-ESM-MR and ACCESS1-3 are better at reproducing interannual mode of EASM,while HadGEM2-AO,GFDL-ESM2G,ACCESS1-0,INMCM4 and GFDL-ESM2M are better for the decadal mode.Five optimal models for interannual and decadal mode of EASM are ensemble averaged and recorded as MME1 and MME2,respectively.The simulation skill scores of MME1 and MME2 are higher than those of all single models,which fully illustrates the importance and superiority of model evaluation and selection.?5?The ability of CMIP5 models to simulate the relationship between the dominant modes of EASM and the underlying thermal anomalies needs to be improved.Overall,the simulation effect of models on the thermal driving factors of the interannual mode is better than that of the decadal mode.About 10 models can simulate the effect of SSTAs from tropical Indian Ocean to Philippine Basin on interannual model of EASM,but only 4 models can reproduce the negative SSTAs in the tropical Atlantic Ocean.Six models can simulate the effect of the thermal anomalies in Lake Baikal on decadal model of the EASM,but the intensity and scope are underestimated.Furthermore,the use of MME1 and MME2 do not significantly enhance the ability to simulate the thermal driving factors of the dominant modes.