| description abstract | Short equatorially trapped Rossby waves are usually thought to he trapped and dissipated nea the region of creation because of their relatively slow group speeds and because of the relatively low amplitude basic state within which they reside. Only long Rossby waves, with stronger group speeds, are assumed to move far away from the forcing area producing remote effects or teleconnections. These restrictions on the regions of influence of Rossby waves are only valid in a motionless basic state or in a basic state that is constant with longitude. If the basic state changes in the zonal direction, even short waves have considerable remote effects. Using ray-tracing techniques and nonlinear numerical models, the impact of a zonally varying basic state on the characteristics of equatorial modes is investigated. The original low-latitude energy accumulation zone theory of Webster and Chang is extended to include the complete family of tropical waves. In equatorial regions, the majority of Rossby waves are longitudinally trapped in regions where the stretching deformation of the background flow is negative. Most of the Rossby packet will reach the energy accumulation area from the east. This kind of wave action flux accumulation (i.e., from the east) is referred to as ?forward accumulation.? However, some shortwaves of the packet will propagate into the accumulation region from the west. This reverse propagation into an accumulation zone is referred to as ?backward accumulation.? Mixed Rossby-gravity waves are also considered. Ray tracing technics indicate that the mixed wave is less likely to be longitudinally trapped. However, if the wave is trapped, the energy accumulation area is generally located in the same place as that of Rossby wave and that the energy is accumulated through backward processes. A nonlinear global spectral model is used to cheek the WKB approximations used in the ray tracing. The results of this study suggest that the ubiquity of the longwave approximation for equatorial modes should be questioned. As the basic state modifies the scale of the mode such that initially very long modes may evolve to shorter scales during propagation through the longitudinally varying flow (and vice versa for initially short modes), the approximation appears questionable in any region where the magnitude of the basic state is similar to the group speeds of the mode, the basic state changes sign, and the basic state possesses a significant stretching deformation. With the longwave approximation, not only does the mixed Rossby-gravity wave disappear but the dispersive Rossby modes are rendered effectively nondispersive. Furthermore, the accumulation property of equatorial waves is also eliminated. | |
