Simulation of the madden-julian oscillation using general circulation models

Abstract

In this chapter, we summarize our experiences with respect to the representation of the Madden- Julian oscillation (MJO) in general circulation models (GCMs). Specifically, we focus on diagnostics that have been used in evaluating GCM simulations, the empirical relationships that have been reported between GCM configurations and MJO simulation skill, and the theoretical frameworks that have been used to interpret the empirical relationships. Popular MJO simulation diagnostics that capture salient features of the MJO were recently standardized by the U.S. CLIVAR MJO Working Group, which make tracking of improvements possible across different models and model generations. Simulation of the MJO is sensitive in particular to the details of the cumulus parameterization used in GCMs. Inhibiting too-frequent activation of vigorous deep convection, which can be achieved a number of ways, generally leads to enhanced MJO variability. Recent MJO theoretical developments have led to new frameworks in which to interpret MJO modeling results. Different, but not mutually exclusive, theoretical frameworks have been applied in different studies, reflecting the lack of a single, satisfactory theory of the MJO. Possible future directions for further improvement of the MJO in GCMs are also discussed.