Comparison of the simulated outcomes of aerosol–cloud interaction by a meteorological model with and without an interactive chemistry module

Abstract

With the general rise of computational capacities and the continued efforts to improve computational efficiencies, more and more studies have been utilizing state-of-the-art atmospheric models that enable cloud-resolving simulations over a global domain. Microphysical processes inside clouds, however, are on a scale much smaller than that of a cloud itself, and therefore resolving clouds in a model is not equivalent to resolving cloud microphysical processes. When aerosol–cloud interaction (ACI) is studied, chemistry models enable the prognostic calculations for chemical species, including aerosols, which can perturb cloud microphysics and eventually impact clouds and climate. The large drawback of these models is the high computational cost required for tracking chemical species in space and time that may not be affordable in some studies. As a result, some studies have used non-chemistry models with prescribed cloud droplet number concentrations n_c and compared multiple simulations with different n_c to assess the impacts of varying aerosol concentrations on clouds. In this study we assess whether the same or similar ACI can be simulated when the aerosol number is increased in a chemistry model and when n_c is raised in a non-chemistry model. A case study has been conducted over the Maritime Continent in September 2015 when an extremely large number of aerosols were observed due to fires under a dry condition brought by a strong El Niño. The comparison of the simulations by the chemistry and non-chemistry models shows that the aerosol-driven enhancement of rainfall in the chemistry simulations was not present in the non-chemistry simulations, even with prescribed n_c with certain spatial variation based on the chemistry runs. Therefore, simulated ACI may largely differ depending on the manner by which an increase or decrease of aerosols is represented in a model. The result suggests the further need for powerful computational capabilities and the pursuit of a rigorous means to incorporate aerosol species in a non-chemistry model.