Convectively Coupled Kelvin Waves in Current and Future Climates

Abstract

Convectively coupled Kelvin waves (KWs) drive tropical precipitation variability on the subseasonal timescale. The KWs influence extreme rainfall and drought events, tropical cyclogenesis, and the Southeast Asian summer monsoon. Despite their profound impacts, the response of KWs in a warmer climate remains an open question due to an incomplete understanding of their dynamics. This study aims to better understand the maintenance and propagation mechanisms of KWs, as well as their response to global surface temperature changes, using observations and aquaplanet simulations. Previous studies suggest that KWs are maintained by two distinct processes: (1) the internal thermodynamic feedback between KW diabatic heating and temperature anomalies and (2) the external momentum forcing that midlatitude Rossby waves exert on KW zonal wind anomalies. However, the relative importance of the two processes is unknown. The relative importance of the two processes to KW maintenance is quantified by comparing the growth rate of eddy available potential energy (EAPE) and eddy kinetic energy (EKE) within KWs using satellite and reanalysis data. Results show that internal thermodynamic feedback is the dominant mechanism that maintains KWs in all regions and seasons. Among all regions, the external forcing is the least important over the Indian Ocean, possibly associated with the highest sea surface temperature and weakest westerly in the upper troposphere, enhancing diabatic heating and inhibiting midlatitude waves approaching the tropics. Results highlight the importance of understanding the convection-wave interaction within KWs in observations and numerical simulations. To investigate changes in the maintenance and propagation of KWs in response to surface warming, I use a set of three aquaplanet simulations with varying sea surface temperature boundary conditions, representing the current climate, a warmer (+4K), and a cooler (-4K) climate. Results show that KWs accelerate at the rate of about 7.1%/K, and their amplitudes decrease by 4.7%/K. The dampening of KWs with warming is associated with a weakening of the internal thermodynamic feedback between diabatic heating and temperature anomalies that generates KW EAPE. The phase speed of KWs closely matches that of the second baroclinic mode KW in -4K, while the phase speed of KWs is approximately that of the first baroclinic mode KW in +4K. Meanwhile, the coupling between the two baroclinic modes weakens with warming. I hypothesize that in -4K, as the first and second baroclinic modes are strongly coupled, KWs destabilize by positive EAPE generation within the second baroclinic mode, and they propagate slower following the second baroclinic mode KW phase speed. In +4K, as the first and second baroclinic modes decouple, KWs are damped by negative EAPE generation within the first baroclinic mode, and they propagate faster following the first baroclinic mode KW phase speed. Then, I examine what controls the coupling between the two modes and why the coupling weakens with warming. While previous simple models of KWs proposed several mechanisms for the coupling, none can fully explain my simulation results. My simulation results suggest the coupling mechanism involves the second baroclinic mode temperature anomalies that modulate deep convection by moistening the lower troposphere via vertical advection of moisture. As the surface warms, the second baroclinic mode diabatic heating anomalies have a weaker magnitude above the top of the boundary layer. With a weaker temperature perturbation in the lower free troposphere associated with it, the second baroclinic mode temperature anomalies appear to be less effective at triggering deep convection and weaken the coupling between the two baroclinic modes. The change in the structure of the second baroclinic mode temperature anomalies with warming is associated with the changes in the mean state temperature profile, especially the rise of the melting level. As the melting level rises, the peak of the second baroclinic mode cooling, which is contributed by melting, is located higher. Thus, the second baroclinic mode diabatic heating anomalies have a weaker magnitude above the top of the boundary layer. These results highlight that the lower tropospheric moistening is crucial to the coupling mechanism and, hence, the maintenance mechanism of KWs. Results suggest that accurate simulation of the mean state changes and the convection-wave coupling is the key to a reliable simulation of KWs.