The MJO skeleton model with observation-based background state and forcing

The MJO skeleton model, a nonlinear oscillator model, has previously been shown to have solutions that exhibit several features in common with observed MJOs. In prior studies, model solutions have been found using mathematically simple (constant or sinusoidal) identical moistening and radiative forcing functions. Here we investigate whether this model can also produce realistic regional variability. To do this, observation-based forcing functions are prescribed for latent heat flux and radiative cooling, and model solutions–both linear and nonlinear–are studied. In the stochastic nonlinear model, solutions reproduce the climatological mean convective activity well and the climatological variance reasonably well for such a simple model. In the linearized model, the solutions are found to contain additional structure including a realistic wave envelope of convective activity centered over the warm pool. These linear solutions are then used to identify and compare significant MJO activity in both reanalysis data and stochastic model solutions. Additional potential uses for these analytical and numerical solutions are discussed.