An Ensemble Approach to Examining the Role of Cloud Condensation Nuclei on the Evolution of Idealized Tropical Cyclones

The Saharan air layer (SAL) is a warm, dry, dusty layer of air that frequently resides over the tropical Atlantic Ocean. In the SAL, cloud condensation nuclei (CCN) concentrations may be enhanced. Some recent modeling studies claim increases in CCN concentration are correlated with decreases in tropical cyclone (TC) intensity, whereas others suggest that there is a non-monotonic relation between CCN and TC intensity. Ultimately, the degree to which CCN in the SAL play a role in TC development and evolution remains unknown.

In this study, an ensemble of simulations is conducted using the Regional Atmospheric Modeling System version 4.3 (RAMS v4.3) to determine whether increases in CCN have a predictable impact on TC evolution when accounting for non-linear amplification of initial noise in model fields. An initial horizontally homogenous 3-D vertical profile of CCN with varying concentrations of 100, 101, 1000, and 2000 cm^-3 between 1 and 5 km is incorporated into a series of simulations where the characteristics of the initial vortex used to initiate convection are perturbed. In particular, the ensemble members consist of small changes in the radius of maximum winds and depth of the initial vortex, the strength of the warm bubble used to initiate convection and the temperature and humidity of the Jordan sounding that is representative of conditions found in the SAL. Simulations with concentrations of 101 cm^-3 are included to test the impact of small perturbations in CCN on simulated storm properties. The resulting 96 hour evolution of maximum surface wind speeds, wind fields, minimum sea level pressure, and distributions of simulated radar reflectivity, diabatic and latent heating rates, vertical velocities, hydrometeor properties, and surface precipitation are examined in order to determine whether perturbations in initial conditions mask aerosol effects on TCs. The relationship of simulated properties in the eye wall and spiral rainbands is also discussed. Physical explanations for the trends in simulated properties are presented.