Based on the reanalysis data and FGOALS-g3, this study explores the contributions of air-sea interactions to the simulation of the East Asian Summer Monsoon (EASM). For climatological mean, the Eastward shift of the Summer Monsoon Rain Belt in atmospheric general circulation models (AGCM) results in dry bias over the East Asian monsoon region. While coupled general circulation models (CGCM) present a proper position of the Summer Monsoon Rain Belt, rain in the East Asian monsoon region is weak. Due to the absence of air-sea interaction, the fake convective responses over the Northwest Pacific region lead to an eastward shift of Subtropical High, which is the main reason for the bias of AGCM in simulating the EASM precipitation. The dry humidity over land is not negligible and contributes 70% of the dry error with the help of the circulation bias. The position bias of Subtropical High and Summer Monsoon Rain Belt is significantly reduced in CGCM which reduces the dry bias by 36%. Less evaporation and less zonal water vapor transport in CGCM simulation together contribute more than 70 % dry bias. For responses to pre-winter El Ni?o, CGCM can reproduce the Western North Pacific Anomalous Anticyclone (WNPAC) and the corresponding dipole precipitation anomaly distribution in the El Ni?o decaying summer. It can also reproduce the Indo-western Pacific Ocean capacitor (IPOC) maintaining WNPAC. In AGCM simulation, convection in the Northwest Pacific, the Bay of Bengal, and the oceans surrounding the Indian Peninsula is too sensitive to SST in the El Ni?o decaying summer. On one hand, this results in strong convective mass transport in the Northwest Pacific, which inhibits the establishment of WNPAC to some extent. For another hand, strong convection in the North Indian Ocean can inhabit convection in the Indian Ocean, resulting in the inability to simulate IPOC mode. Therefore, the lack of air-sea interaction is the main reason for why AGCM cannot reproduce the two-season lag response of EASM to major El Ni?o.