摘要:
The magnitude of El Niño determines the level of its global impact. Yet, how well our state-of-the-art models simulate this key aspect of El Niño is not well documented. Previous studies tend to ignore ENSO asymmetry and equate the variance of ENSO to the magnitude of El Niño. Moreover, previous evaluations are more focused on the surface manifestation of El Niño. Here, we quantify the magnitudes of El Niño and La Niña separately, both at the surface and subsurface levels. At the surface, we find that while the magnitude of La Niña events in most models is generally stronger than observed, the magnitude of El Niño is more diverse to observations. In fact, in many models, El Niño is weaker than observed. This bias in the magnitude of El Niño is more pronounced in the subsurface. We attribute this weakness in the subsurface to the generally weaker coupling strength and the apparent stronger ENSO at the surface to a lack of sufficiently strong negative feedback from the surface heat flux in the models. When normalized by the variance of ENSO, the lack of exceptionally strong El Niño events in the models is more common and pronounced. The consequences of a generally weaker El Niño in the models are discussed.