Using the CSEOF technique as a detecting tool for intermittent oscillations, we could successfully extract the eastward moving large scale convective system of dipole structure. The accompanying circulation showed the prominent first baroclinic vertical structure, which is coherently related with the heat source. As the heat source moved eastward from the Indian Ocean, the regressed stream function also moved with the heat source. The phase difference between the heat source and the circulation was almost the same during life cycle of the convective system. Motivated by the above mentioned phase difference, we compared the analysis result with the Gill type solution forced by the same heat source. In the life cycle of OLR extracted by CSEOF, the Gill model was forced by the heating estimated from the OLR pattern. Although Gill model does not have any time dependency, the model response was well matched with the analysis result as described above. Thus, we could interpret that the local response confined in the tropics is due to the almost instantaneous adjustment of the atmosphere on the traveling heat source.
In the second phase of this study, we examined the circulation pattern deviated from the Gill type that is directly forced by traveling heat source. To assess the mechanism of the circulaton, we expanded our analysis domain to the midlatitude. The circulation was characterized by varying in harmony with the tropical heat source. The Positive and negative phase of PNA pattern was observed sequentially during the life cycle of MJO, as noted by previous researches. Thus, the midlatitude circulation extracted by CSEOF seemed to have close connection with the tropical traveling heat forcing. We examined the hypothesis that the midlatitude circulation is indirectly forced by the tropical heat source. For clarifying this implication, we conducted budget analysis using regressed variables such as stream function and velocity potential with barotropic vorticity equation. Constituting Rossby wave source terms were calculated. The primary balance can be described as that subtropical vortex stretching compensates the vorticity advection by the nondivergent flow. The absolute vorticity advection by divergent wind always observed near the southern flank of vortex stretching. Thus, we may conclude that divergent wind transports absolute vorticity northward, which resulted in the subtropical vorticity source. This source might be in charge of midlatitude intraseasonal variability.
Supplementary URL: