Simulation of the Madden–Julian Oscillation in a Coupled General Circulation Model. Part I: Comparison with Observations and an Atmosphere-Only GCM

Abstract

The simulation of the Madden?Julian oscillation (MJO) has become something of a benchmark test for the performance of GCMs in the Tropics over recent years. Many atmospheric GCMs have been shown to reproduce some aspects of the MJO but have had problems representing its amplitude, propagation speed, and seasonality. Recent observational and modeling studies have suggested that the MJO is, at least to some extent, a coupled phenomenon. Thus with the complex interactions between convection and large-scale dynamics, together with the interactions between the sea surface and boundary layer, the MJO provides a rigorous test for many aspects of a GCM formulation. In this study, the ability of an atmosphere?ocean coupled global climate model to represent various aspects of the MJO will be examined, and compared with the performance of the atmosphere-only component of the same model forced with slowly varying sea surface temperatures. One impact of coupling this GCM to an interactive ocean is to improve the eastward propagation of convection across the Indian Ocean. Surface flux anomalies associated with the MJO are in reasonable agreement with observations, although the resulting SST variability is found to be slightly weaker than observed. There is no propagation of convection into the west Pacific in the coupled model. It is proposed that this deficiency is due to errors in the basic state of the coupled model, in particular the lack of low-level westerly winds over the west Pacific.