Comparison of different model approaches for the simulation of multiphase processes

摘要

Cloud-chemistry models are developed intensively with increasing complexity, leading to new knowledge and offering new possibilities to understand the physico-chemical processes taking place in the atmosphere. Intercomparing such detailed models is the way to test the robustness and reliability of their parameterizations and numerical schemes. The present study involves newly developed parcel models treating microphysics and chemistry with equal rigor. The description of both kinds of processes is given for a size-resolved particle/droplet spectrum. Three different types of models are compared. In the SPACCIM approach, one- and two-dimensional particle/drop microphysical schemes are used in a time-splitting setup between chemistry and microphysics. The GALERKIN model employs a one-dimensional scheme in a fully coupled setup. For each of the three types, "fixed bin" and "moving bin" approaches are implemented. A comparison between "fixed" and "moving bin" approaches makes sense only for scenarios without coagulation and breakup. The paper focuses on the effects of different microphysical and numerical approaches on the multiphase chemistry. The resulting changes in the particle/droplet composition feed back on cloud microphysics. Substantiated conclusions can only be derived if these effects are studied for a wide range of cases. Thus, the simulations are performed for three chemical reaction mechanisms of different complexity and four scenarios derived from field measurements. The interaction between numerical schemes, microphysics and multiphase chemistry is discussed. Mostly, the results of the participating models agree in an appreciable way. Observable differences are noticed between the "moving bin" approach and models using fixed grids for the discretization of the particle/droplet spectrum. Furthermore, the initial aerosol composition influences the fate of chemical species as well as the behavior of the numerical solver in a substantial way.