In our study, we mainly focus on a series of convective events occurred near the end of the observation period with strong emphasis on scale interaction process. Our observation period (28 October-21 November) could be divided into a dry phase before and a wet phase after 16 November. Almost all organized deep convection was observed during the wet phase. In situ data including sounding and Doppler radar together with the JCDAM reanalysis are used to document what the wet phase look like. A careful separation of disturbances into several types of convectively coupled equatorial waves shows what was observed is the moist Kelvin wave encompassing westward propagating inertio-gravity (WIG) waves followed by the MJO convection. The moist Kelvin wave provides low level convergence to the east of its major convective body by about several thousand kilometers. About two WIG waves are included and enhanced in this convergence zone; one is around the eastern tip of the convergence zone and the other is around the major convective body of the moist Kelvin wave. The WIG wave located to the east moistens middle troposphere, which makes the environment more preferable for deep convection, namely preconditioning. The WIG wave located to the center is responsible for taking over the major convective body of the moist Kelvin wave.
Although unfortunately we missed to observe the MJO itself, from the similarity between the MJO and moist Kelvin wave in terms of multi-scale structure it is beyond doubt that the result obtained here is useful to improve our knowledge of the scale interaction of not only the moist Kelvin wave but also the MJO with the WIG waves. A numerical experiment using a nonhydrostatic mesoscale model is planned to further elucidate the scale interaction process between the moist Kelvin wave and WIG waves.