Abstract
This study provides an explanation for the origin of the tropical intraseasonal (40–50 day) oscillation (TIO) based on a simple generalization of Gill's linear analytic model for tropical large-scale heat-induced circulation. The solution, which compares favorably with observations, contains a convective region that excites an eastward-moving Kelvin wave and a westward-moving Rossby wave. The significance of the Rossby wave, not previously emphasized, is clearly revealed. The entire system moves eastward as a response to the circulation it excites at a speed at which the latent heat energy in the tropics is best extracted. Thus, the TIO is viewed as an intrinsic instability. Its speed is related to the vertical heating profile and is a decreasing function of both dissipation and the zonal size of the convective region. This speed is a weighted mean of the speed of the Kelvin wave and that of the Rossby wave. Previous studies have erroneously equated the speed of the TIO with that of the Kelvin wave. This study also demonstrates that classification of the TIO as a wavenumber 1 phenomenon is not advisable.