Large-scale intraseasonal moisture convergence due to warm SST before convection

Observations from the R/V Revelle during DYNAMO test theories of the MJO. First, column integrated moisture (6 cm liquid water equivalent in the mean) varied intraseasonally with an amplitude of ±2 cm. High column moisture was related to but not a sufficient condition for deep convection. Nor did precipitation simply discharge moisture anomalies. Column integrated moisture increased on the days with highest precipitation at the beginning of the convective phase, and decreased during weaker trailing precipitation later in the convective phase. Second, evaporation (110 W m^-2 in the mean) increased to 150 W m^-2, in the convective phase of the MJO, explaining about 10% of the precipitation during the convective phase. Consistent with previous studies from TOGA-COARE and DYNAMO, the magnitude of evaporation anomalies is likely not sufficient by itself to overcome drying processes during the convective phase.

Noting the inadequacy of local surface evaporation anomalies to support MJO moisture anomalies, we propose an alternative model of air-sea interaction for the MJO. We assume that the boundary layer is near thermal equilibrium with the sea surface temperature (SST), which increases during the suppressed phase and decreases in the convective phase of the MJO. Peak SST occurs several days before the convective phase, partly due to diurnal warm layers. The warm SST anomalies induce hydrostatic moisture convergence of ~50 W m^-2 in the boundary layer in advance of the convective phase. Winds and horizontal moisture convergence in the boundary layer are estimated with a hydrostatic model.