To begin or not to begin? A case study on the MJO initiation problem

Mechanisms for convective initiation of the Madden-Julian oscillation (MJO) remain poorly understood. During recent years, < 50 % of large-scale convectively active episodes over the tropical Indian Ocean have led to MJO initiation. This study explores the structure and evolution of precipitation, diabatic heating, and potential vorticity (PV) that might be used to tell whether an MJO event will be initiated once such a convection episode occurs. Three different cases are studied. As convection becomes active in a large area over the tropical Indian Ocean, early signs favorable for MJO initiation are apparent: a persistent basin-scale coverage in the zonal direction by positive anomalies in precipitation and diabatic heating (in a swallowtail pattern), a persistent vertical dipole of PV generation with cyclonic (anticyclonic) PV generation in the lower (upper) troposphere covering a zonally extended area, and a cyclonic PV anomaly in the midtroposphere with a cyclonic PV pair straddling the equator immediately west of the diabatic heating center. All these signs are robust in the MJO composite but rarely occur all together in a given MJO case. The likelihood of an MJO event following a convective episode over the tropical Indian Ocean depends on how many of these signs occur and how persistent they are. While a preexisting MJO signal is neither a necessary nor a sufficient sign for MJO initiation, an active convective episode over the tropical Indian Ocean is necessary but insufficient for MJO initiation. MJO initiation depends on detailed convective behaviors over the tropical Indian Ocean.