Diversity of the Pacific-Japan Pattern among CMIP5 Models: Role of SST Anomalies and Atmospheric Mean Flow

This study investigates the reproducibility of the spatial structure and amplitude of the observed Pacific-Japan (PJ) pattern in the phase 5 of the Coupled Model Intercomparison Project (CMIP5) models. In particular, the role of sea surface temperature anomalies (SSTAs) and atmospheric mean flow in the diverse reproducibility of the PJ pattern among models is investigated. Based on the pattern correlation between simulated and observed PJ patterns, models are categorized into high and low correlation groups, referred to as HCG and LCG, respectively. The observed cold SSTAs in the western North Pacific (WNP) and equatorial central Pacific, organized convection and precipitation anomalies, and Rossby wave response are reproduced well in HCG models, whereas these features are not present in LCG models. The summer SSTAs are closely tied to the preceding El Nino-Southern Oscillation and its temporal evolution in the tropical Indo-Pacific Ocean in both observations and models, but the SSTAs in the Indian Ocean are weak in both HCG and LCG, implying a weak Indian Ocean capacitor effect. As a result, the reproducibility of the amplitude of the WNP center of the PJ pattern is mainly modulated by the SSTAs and local air-sea feedback over the WNP in the models. On the other hand, a model with stronger climatological southerly along the coast of East Asia tends to produce more realistic amplitude of the midlatitude center of the PJ pattern with clearer poleward wave-activity fluxes due to more efficient local barotropic energy conversion from the mean flow.