Modeling aerosol activation in a tropical, orographic, island setting: Sensitivity tests and comparison with observations

Abstract

The aerosol, updraft and cloud droplet observations from the 2011 Dominica Experiment (DOMEX) field campaign provide an interesting opportunity to investigate the process of cloud droplet activation in a tropical, orographic, convective setting. This study involves adiabatic parcel model simulations with a state-of-the-art parameterization of droplet activation, which we run with aerosol size distributions and updraft velocities based on DOMEX data. We compare the cloud droplet concentrations predicted by the parameterization with the observations from DOMEX, and run various sensitivity tests to changes in model inputs on the order of their uncertainty, in order to gain insights into what factors are most important in determining the aerosol activation fraction in this setting. Our control simulations overestimated the observed droplet concentrations, especially for the days with strong trade winds, but in most cases these discrepancies could be eliminated by realistic changes in our assumptions. The remaining error could be the result of entrainment of sub-saturated air, precipitation, or advection of pre-existing clouds from upwind. We found strong sensitivities to the mean updraft velocity and to the size distribution and composition of particles in the Aitken mode, the smallest mode including particles below 100nm. The Aitken mode accounted for 42% to 68% of the simulated droplet concentration in our control simulations, and simulations excluding the Aitken mode underestimated the observed droplet concentrations under realistic assumptions. Droplets from the Aitken mode dominated the changes in the simulated droplet concentrations in our sensitivity tests. The precision of our simulations, and our ability to constrain the role of the Aitken mode, were limited by our lack of knowledge of the composition and size distribution of Aitken mode particles, highlighting the importance of measuring these variables in field campaigns in similar settings.