The Impact of Indian Ocean Mean-State Biases in Climate Models on the Representation of the East African Short Rains

Abstract

The role of the Indian Ocean dipole (IOD) in controlling interannual variability in the East African short rains, from October to December, is examined in state-of-the-art models and in detail in one particular climate model. In observations, a wet short-rainy season is associated with the positive phase of the IOD and anomalous easterly low-level flow across the equatorial Indian Ocean. A model’s ability to capture the teleconnection to the positive IOD is closely related to its representation of the mean state. During the short-rains season, the observed low-level wind in the equatorial Indian Ocean is westerly. However, half of the models analyzed exhibit mean-state easterlies across the entire basin. Specifically, those models that exhibit mean-state low-level equatorial easterlies in the Indian Ocean, rather than the observed westerlies, are unable to capture the latitudinal structure of moisture advection into East Africa during a positive IOD. Furthermore, the associated anomalous easterly surface wind stress causes upwelling in the eastern Indian Ocean. This upwelling draws up cool subsurface waters, enhancing the zonal sea surface temperature gradient between west and east and strengthening the positive IOD pattern, further amplifying the easterly wind stress. This positive Bjerknes coupled feedback is stronger in easterly mean-state models, resulting in a wetter East African short-rain precipitation bias in those models.