Aerosol indirect forcing of the trimodal distribution of tropical convection within a radiative convective equilibrium framework

Tropical convection tends to be distributed in a trimodal manner, being made up of a shallow cloud mode, a congestus mode and a deep convective mode. Much is still not understood about these three modes of convection including how these modes interact, the roles that they serve in the global water and energy balance, and how these modes may change when perturbed by external effects such as aerosol indirect forcing. The goal of the research to be presented here is to investigate the impacts of aerosol indirect forcing on the properties, organization and interaction of these three modes of tropical convection. As the tropical atmosphere is never far from a state of radiative convective equilibrium (RCE), conducting cloud-resolving model (CRM) studies under a RCE framework has proven highly successful in a number of experiments focusing on the feedbacks between radiation, clouds, water vapor and convection in the tropics. Numerous numerical simulations have thus been conducted using the Regional Atmospheric Modeling System (RAMS) within such a RCE framework in order to achieve the stated goal.

RAMS is a sophisticated CRM that allows for the prognosis of aerosol concentrations. Two-dimensional simulations have been performed using a model grid that spans approximately 10,000 km in the zonal direction, with a horizontal grid spacing of 1km, variable grid spacing in the vertical, and periodic lateral boundary conditions. The lower boundary is an oceanic boundary with fixed sea surface temperature. The model is initialized using the 00 GMT 5 December 1992 TOGA COARE sounding and convection is initiated by randomized perturbations to the potential temperature. The model is first run until radiative convective equilibrium is reached which takes approximately 50 simulation days. Numerous sensitivity tests are then conducted in which a layer of aerosol that can potentially serve as cloud condensation nuclei (CCN) and/or ice nuclei (IN) is introduced between 2 and 4 km AGL. This setup is representative of a Saharan dust event over the Atlantic Ocean. The model is restarted at day 50 with this aerosol layer and is run for another 50 days. The amount of aerosol available for activation is progressively increased from one sensitivity experiment to the next, thus representing a range of aerosol conditions from pristine to polluted. The use of such large-domain, high-resolution and temporally long simulations allows for the investigation of aerosol indirect forcing on the wide range of different cloud types and systems that develop under the variety of environments that evolve.

Initial results from these experiments demonstrate that the precipitation response to higher CCN concentrations is dependent on the storm type; that heavy rainfall producing systems become more intense but less frequent as CCN is increased; that enhanced CCN concentrations result in stronger updrafts that cover greater fractions of the domain; that the ratio of liquid water to ice mass is sensitive to increased CCN concentrations; and that greater concentrations of CCN produce a mixed response in the frequencies of the three modes of convection. The influence of aerosol on the relative contributions of the three modes to the liquid water path, ice water path and surface precipitation have also been examined. Finally, the impacts of variations in aerosol concentrations on all the terms contributing to the water vapor budget have also been assessed, and these results will also be presented.