The evolution of El Niño–Southern Oscillation (ENSO)-monsoon relationship have been studied using a suite of 55-yr experiments with a 62-wavenumber, 64-level general circulation model. The "pacemaker" experiment, in which tropical Pacific SST is prescribed from observations but coupled air-sea feedbacks are maintained in the other ocean basins, is performed for the 1950–2004 period. The “control” run by using climatological values instead of coupling with the slab ocean model is also performed. Four independent integrations have been conducted for each of these two forcing scenarios. CFS CGCM free long run during 52-year is also compared as the “coupled” experiment.

The relationship between ENSO and Asian monsoon is not only a good example of air-sea coupled co-variability, but also an interesting subject for this model because CFS CGCM long run cannot mimic this relationship very well. CGCM shows weak relationship with Indian monsoon with insignificant confidence level and delayed signal with western North Pacific monsoon different from observation. Surprisingly, simulated relationship in pacemaker is reasonably close to observed. Analysis of dominant mode of monsoon field associated with ENSO also shows good accordance with observed. This improvement in pacemaker comparing to CGCM shows that simple systems with the only local air-sea feedback are capable of greater realism than sophisticated coupled models, since the CGCM has significant errors in the ocean dynamics and Walker circulation related with ENSO. The control run without coupled ocean also simulates it less well.

In pacemaker, decadal change of ENSO-Indian monsoon relationship is not shown for 55-year period, while clearly shown for western North Pacific. It suggests decadal change of eastern Pacific ENSO anomalies strongly linked with that of western North Pacific monsoon. Based on this result, the cause of decadal change of monsoon is investigated.