Abstract: Observations show evident negative correlation between summer sea surface temperature (SST) and precipitation anomalies over the northwestern Pacific (NWP), which is due to the maintenance of the NWP anomalous anticyclone from the El Niño peak winter to subsequent summer. This process has been a useful metric for assessing the capability of climate models. Based on the near-term climate prediction system of the Institute of Atmospheric Physics (IAP-DecPreS), this study evaluates the result of the ocean data assimilation experiment using the coupled model FGOALS-s2 to simulate the air-sea interaction over the NWP through comparing the result with observations and standalone AGCM (atmospheric general circulation model) simulation. The coupled data assimilation system performs better than AGCM and reasonably reproduces the negative correlation between summer SST and precipitation anomalies over the NWP, although the area with negative correlation coefficient is smaller than the observation, especially in August when the NWP anticyclone is weaker than that in the observation and shifts eastward. Further analysis indicates that the bias is associated with the southward shift and weakening of positive precipitation anomalies over the equatorial eastern Indian Ocean in the El Niño decaying summer. The standalone AGCM experiments fails in simulating the negative correlation between summer SST and precipitation anomalies over the NWP, which is attributed to the absence of atmospheric forcing. Our analysis demonstrates that the ocean assimilation system performs better than the AMIP experiment in simulating the air-sea interaction over the NWP, and the bias in the NWP anticyclone simulation in August is due to the bias in precipitation anomalies over the equatorial eastern Indian Ocean.