Governing thermodynamic ttructure of Tropical East Pacific and African Easterly Waves and their representation in CMIP6 models

Abstract

Easterly Waves (EWs) are convectively coupled systems of alternating high and low pressure that propagate westward in the Tropics with a period of 3-6 days and horizontal scale of ∼3000 km. They are most commonly observed in the western hemisphere over the east Pacific Ocean (PEWs) and Africa and the east Atlantic Ocean (AEWs). While significant progress has been achieved in understanding PEWs and AEWs, a thorough comparison between the waves as well as their representation in global climate models (GCMs) remains elusive. The thermodynamic processes associated with convection in AEWs, and PEWs are examined on the basis of empirical orthogonal functions (EOFs) and a plume buoyancy framework. Plume buoyancy is found to be highly correlated with rainfall in both AEWs and PEWs. Close inspection of the contribution of moisture and temperature to plume buoyancy reveals that temperature and moisture contribute roughly equally to the buoyancy in AEWs, while moisture dominates the distribution of buoyancy in PEWs. The results of this study indicate that PEWs are moisture modes:waves in which water vapor plays a dominant role in their thermodynamics. AEWs,on the other hand, are mixed waves in which temperature and moisture play similar roles in their thermodynamics. The representation of these governing thermodynamic structures is further examined in historical Coupled Model Intercomparison Project phase 6 (CMIP6) models. Simulations are compared to data from the Fifth Reanalysis from the European Centre for Medium-Range Weather Forecasts (ERA5). Two representation skill metrics are employed to select models with accurate PEW and AEW representation,the (1) wave relative amplitude and (2) spatial coherence obtained from the pattern correlations of lag-regressed anomalous precipitation between ERA5 and CMIP6 models.A large spread in PEW and AEW representations is observed. Results suggest thatthe spread in representation skill may be tied to the misrepresentation of convection sensitivity to column moisture, the horizontal distribution of mean column moisture, and precipitation. Furthermore, examination of model diagnostics reveal that both waves grow from moisture-vortex instability, even in the presence of strong meridional temperature gradients. These results underscore the relevance of the accurate representation of moist processes for the accurate representation of PEWs and AEWs in CMIP6 Models.