MATRIX (Multiconfiguration Aerosol TRacker of mIXing state): an aerosol microphysical module for global atmospheric models

Abstract. A new aerosol microphysical module MATRIX, the Multiconfiguration Aerosol TRacker of mIXing state, and its application in the Goddard Institute for Space Studies (GISS) climate model (ModelE) are described. This module, which is based on the quadrature method of moments (QMOM), represents nucleation, condensation, coagulation, internal and external mixing, and cloud-drop activation and provides aerosol particle mass and number concentration and particle size information for up to 16 mixed-mode aerosol populations. Internal and external mixing among aerosol components sulfate, nitrate, ammonium, carbonaceous aerosols, dust and sea-salt particles are represented. The solubility of each aerosol population, which is explicitly calculated based on its soluble and insoluble components, enables calculation of the dependence of cloud drop activation on the microphysical characterization of multiple soluble aerosol populations. A detailed model description and results of box-model simulations of various aerosol population configurations are presented. The box model experiments demonstrate the dependence of cloud activating aerosol number concentration on the aerosol population configuration; comparisons to sectional models are quite favorable. MATRIX is incorporated into the GISS climate model and simulations are carried out primarily to assess its performance/efficiency for global-scale atmospheric model application. Simulation results were compared with aircraft and station measurements of aerosol mass and number concentration and particle size to assess the ability of the new method to yield data suitable for such comparison. The model accurately captures the observed size distributions in the Aitken and accumulation modes up to particle diameter 1 μm, in which sulfate, nitrate, black and organic carbon are predominantly located; however the model underestimates coarse-mode number concentration and size, especially in the marine environment. This is more likely due to oversimplifications of the representation of sea salt emissions – sea salt emissions are only calculated for two size classes – than to inherent limitations of MATRIX.