To enhance accurate estimates of Asian monsoon variability by a coupled general circulation model (GCM), the bulk adjustment factors that control latent heat, sensible heat and momentum fluxes are optimized using a 4-dimensional variational data assimilation method. When using the optimized values, a coupled GCM is better able to define the summer monsoon features over Southeast Asia. In particular, the early, rapid onset is realistically simulated around the Indochina Peninsula, which is a key region for the initial stage of the Asian summer monsoon development. The spatial patterns of precipitation rate, wind velocity, and sea surface temperature are successfully reproduced. The optimized values of the bulk adjustment factors for latent heat flux become significantly correlated with the strength of the subgrid-scale disturbances primarily associated with energetic deep convective activities in the tropics, which is one of the major sources of model biases in commonly-used coarse resolution models.