Reexamining the Moisture Mode Theories of the Madden–Julian Oscillation Based on Observational Analyses

Two existing moisture mode theories of the MJO, one emphasizing boundary layer moisture asymmetry (MA) and the other emphasizing column-integrated moist static energy (MSE) tendency asymmetry (TA), were validated with the diagnosis of observational data during 1979–2012. A total of 2343 MJO days are selected. While all these days show a clear phase leading of the boundary layer moisture, 20% of these days do not show a positive column-integrated MSE tendency in front of MJO convection (non-TA). A further MSE budget analysis indicates that the difference between the non-TA composite and the TA composite lies in the zonal extent of anomalously vertical overturning circulation in front of the MJO convection. A background mean precipitation modulation mechanism is proposed to explain the distinctive circulation responses. Dependent on the MJO location, an anomalous Gill response to the heating is greatly modulated by the seasonal mean and ENSO induced precipitation fields. Despite the negative MSE tendency in front of MJO convection in the non-TA group, the system continues moving eastward due to the effect of the boundary layer moistening, which promotes a convectively unstable stratification ahead of MJO convection. The analysis result suggests that the first type of moisture mode theories, the moisture asymmetry mechanism, appears more robust, particularly over the eastern Maritime Continent and western Pacific.

Reexamining the MJO Moisture Mode Theories with Normalized Phase Evolutions

A normalization method is applied to MJO-scale precipitation and column integrated moist static energy (MSE) anomalies to clearly illustrate the phase evolution of MJO. It is found that the MJO peak phases do not move smoothly, rather they jump from the original convective region to a new location to its east. Such a discontinuous phase evolution is related to the emerging and developing of new congestus convection to the east of the preexisting deep convection. While the characteristic length scale of the phase jump depends on a Kelvin wave response, the associated time scale represents the establishment of an unstable stratification in the front due to boundary layer moistening. The combined effect of the aforementioned characteristic length and time scales determines the observed slow eastward phase speed. Such a phase evolution characteristic seems to support the moisture mode theory of the second type that emphasizes the boundary layer moisture asymmetry, because the moisture mode theory of the first type, which emphasizes the moisture or MSE tendency asymmetry, might favor more “smooth” phase propagation. A longitudinal-location-dependent premoistening mechanism is found based on moisture budget analysis. For the MJO in the eastern Indian Ocean, the premoistening in front of the MJO convection arises from vertical advection, whereas for the MJO over the western Pacific Ocean, it is attributed to the surface evaporating process.

Changes in the MJO under Greenhouse Gas–Induced Warming in CMIP5 Models

This study investigates changes to the Madden–Julian oscillation (MJO) in response to greenhouse gas–induced warming during the twenty-first century. Changes in the MJO’s amplitude, phase speed, and zonal scale are examined in five models from phase 5 of the Coupled Model Intercomparison Project (CMIP5) that demonstrate superior MJO characteristics. Under warming, the CMIP5 models exhibit a robust increase in the spectral power of planetary-scale, intraseasonal, eastward-propagating (MJO) precipitation anomalies (~10.9% K−1). The amplification of MJO variability is accompanied by an increase of the spectral power of the corresponding westward-traveling waves at a similar rate. This suggests that enhanced MJO variability in a warmer climate is likely caused by enhanced background tropical precipitation variability, not by changes in the MJO’s stability. All models examined show an increase in the MJO’s phase speed (1.8% K–1–4.5% K−1) and a decrease in the MJO’s zonal wavenumber (1.0% K–1–3.8% K−1). Using a linear moisture mode framework, this study tests the theory-predicted phase speed changes against the simulated phase speed changes. It is found that the MJO’s acceleration in a warmer climate is a result of enhanced horizontal moisture advection by the steepening of the mean meridional moisture gradient and the decrease in zonal wavenumber, which is partially offset by the lengthening of the convective moisture adjustment time scale and the increase in gross dry stability. While the ability of the linear moisture mode framework to explain MJO phase speed changes is model dependent, the theory can accurately predict the phase speed changes in the model ensemble.