Interdecadal Variability of the Meridional Wind across the Eastern Equatorial Pacific and Its Relationship with ENSO

Abstract

Abstract The interdecadal variability of cross-equatorial meridional winds in the eastern Pacific (VEP), which has been shown to correlate with the amplitude of El Niño–Southern Oscillation (ENSO) on the interdecadal time scale, is investigated using long-term observations and 22 models from phase 6 of the Coupled Model Intercomparison Project (CMIP6). Both observations and models exhibit a tight negative correlation between interannual variations of VEP and ENSO, and this relationship is synchronized with the interdecadal variability of VEP. In long-term observations, ENSO amplitude modulation is still seen to be out-of-phase with interdecadal variability of VEP. This relationship, however, is substantially underestimated among CMIP6 models, particularly in the historical simulations. The interdecadal variability of VEP is associated with both the interdecadal Pacific oscillation (IPO) and the Atlantic multidecadal oscillation (AMO) in observations. However, most CMIP6 preindustrial control (PI-control) experiments have no link between the VEP and AMO. In contrast, in historical simulations, the multimodel mean of interdecadal VEP shows a significant fluctuation with AMO around the 1980s, which might be caused by the anthropogenic aerosol forcing. Consequently, the interdecadal variation of the AMO–VEP relationship is likely a response to external forcing while the IPO–VEP relationship is mainly modulated by internal climate variability. Another plausible factor causing the weak AMO–VEP relationship in PI-control runs is the unrealistic relationship between modeled SSTs in the tropical Pacific and North Atlantic. Furthermore, model biases in the tropical Pacific may account for the weak relationship between interdecadal VEP and ENSO amplitude in CMIP6 models.