| description abstract | In this paper, the circulations driven by deep heating and shallow heating are investigated through analytically solving a set of linear equations and examining circulations simulated by a dry primitive equation model. Special emphasis is placed on the low-level mass (moisture) convergence associated with the forced circulation and the maintenance of the shallow and deep heat sources. It is found that the forced circulation driven by shallow heating is more likely to be trapped horizontally near the heating area but relatively extended in the vertical. As a consequence, diabatic heating cannot balance adiabatic cooling due to upward motion. At the levels slightly above the top of the heating, a negative vertical gradient of temperature perturbation appears. For the atmosphere driven by deep heating, however, the temperature perturbation cannot accumulate because the heating signals propagate away very fast, allowing an approximate equilibrium between the convective diabatic heating and adiabatic cooling due to upward motion. The converged moisture associated with circulation driven by shallow heating exceeds the amount needed to maintain the heat source. However, the circulation driven by deep heating does not feed back effectively to the moisture convergence, and thus cannot be self-sustaining. Based on these results, a new mechanism is proposed for the interaction between the large-scale convection and large-scale circulation. The new mechanism states that shallow heating drives a strong low-level moisture convergence so that the system of shallow heating and the forced large-scale circulation is unstable. When the unstable system reaches a certain amplitude, the stable cap layer immediately above the shallow heating erodes, and deep convection arises, which consumes most of the converged moisture at low levels without much feedback to the low-level convergence of moisture. The whole heating circulation system develops and dies; the estimated lifetime of such a system, based on the timescale of adjustment of tropical atmosphere to forcing, is on an intraseasonal timescale. Related observational and modeling evidence that support the new mechanism is discussed. | |
