Coupling Between the First and Second Baroclinic Modes within Convectively Coupled Kelvin Waves

Studies have shown that the representation of convectively coupled equatorial waves (CCEWs) in global weather and climate models are sensitive to whether convection is (partially) resolved or parameterized. This study aims to investigate why storm-resolving global models represent CCEWs differently than the models with parameterized convection.

We performed aquaplanet simulations on different horizonal resolutions (120km, 30km, 15km, and 3km) with prescribed zonally symmetric sea surface temperature mimicking the current climate using the Model for Prediction Across Scales-Atmosphere (MPAS-A).

Results show that tropical precipitation is mostly produced by the convective parameterization scheme in 120 km, 30km, and 15km simulations, while it is mostly from cloud microphysics scheme in the 3km run. Convectively coupled Kelvin waves (CCKWs), which are the dominant mode of tropical precipitation variability in all simulations, are stronger and slower in the 3km run. Stronger CCKWs in the 3km run are consistent with a larger growth rate of the eddy available potential energy (EAPE) within the second baroclinic mode, mostly due to stronger stratiform heating and more in-phase relationship between CCKW heating and temperature anomalies. In the 3km run, stronger stratiform heating also offsets adiabatic cooling within CCKWs to a greater degree, resulting in a weaker effective static stability and hence the lower phase speed of CCKWs. Furthermore, in the 3km run, precipitation is less sensitive to deep convective inhibition and more sensitive to saturation fraction than in the other simulations. Our results suggest that explicitly resolved convection is more effective in producing stronger stratiform heating, likely because the effect of mesoscale organization of convection is more pronounced, which leads to stronger and slower CCKWs.