Triggered Convection, Gravity Waves, and the MJO: A GCM Study

Based on numerical experiments with the shallow water equations, Yang and Ingersoll (2013, JAS, in press, referred to here as YI13) propose that the MJO is an interference pattern of the eastward and westward inertia-gravity (EIG and WIG) waves. The waves are generated by triggered convection events, and the propagation speed is related to the difference in the speeds of the EIG and WIG waves. The horizontal scale is determined by the bandwidth of the excited waves, which is inversely related to the frequency of convective events. Since the EIG and WIG waves are high-frequency waves, the YI13 hypothesis also implies that quasi-equilibrium convective adjustment might not be adequate to simulate the MJO. Instead, triggered convection could be important in simulating the MJO.

In this study, we test the YI13 hypothesis with a 3D moist GCM that does not have MJO signals (Frierson 2007). A simplified Betts-Miller scheme is used in this GCM. We modify the GCM by adding another convective criterion: Convection happens only when the convective inhibition (CIN) falls below a certain threshold. Until that threshold is reached, the atmosphere has time to accumulate convective available potential energy (CAPE), and then the convection is triggered. Figure 1 shows the Hovmoller diagram of precipitation in our 3D simulation. The MJO-like signals are observed, and the propagation speed is ~ 5 m/s. The MJO-like signal also has realistic vertical and horizontal structures. In addition, we find that the inertia-gravity waves are stronger in simulations with improved MJO performance. These results support the YI13 hypothesis. Further tests will be performed and presented.

Fig. 1: Hovmoller diagram of the precipitation in our GCM simulation. The red (blue) represents strong (weak) precipitation.