Convective Coupling and Interannual and Intraseasonal Coupled Variabilities in the Tropics

Abstract

Recent analyses point at close structural similarities and intimate dynamical connections among the members of a wide spectrum of variabilities in the global Tropics. This suggests that these oscillations with widely different timescales may have a common dynamical origin, although diverse dynamical evolution. The present work explores such a dynamical framework for the genesis of tropical variabilities. The highlight of the dynamical framework is a parameterization of air?sea interaction, which the authors call convective coupling. In the proposed parameterization the effect of sea surface temperature (SST) on the atmosphere is only through the modulation of the column precipitation. The changed convective heating modifies the low-level flow, which in turn changes the oceanic variables, thus closing the cycle. Thus the forcing due to convective coupling behaves like a (second order) threshold process, with (SST modulated) convection acting like a switch. Since many of the observed variabilities exhibit a primarily Kelvin wave structure, the implications of convective coupling are investigated for the dynamics of Kelvin waves. To account for the different mean conditions over which the variabilities of different scales occur, two scenarios, characterized by the strengths of certain physical parameters, are considered to represent the interannual and intraseasonal (mean) conditions. Oscillations of appropriate periods appear for each scenario. Two different representations of dynamics of SST are also considered to explore coupled variabilities in ocean basins with different SST dynamics. It is also shown that the convective coupling can support interannual variability in ocean basins with small zonal scale. The dynamical scenario that emerges from this study is that the genesis of a wide spectrum of tropical variabilities is caused by a selective excitation of a convectively coupled Kelvin wave embedded in the time-mean state of the corresponding timescale of oscillation. The postgenesis evolution and structure of each oscillation is then influenced by local spatiotemporal characteristics. The results provide both insight for modeling air?sea interaction in the Tropics and constraints for model diagnostics and evaluation.