A monodisperse aerosol dynamics module, a promising candidate for use in long‐range transport models: Box model tests

[1] Air quality models used as a scientific basis for air pollution strategy development need physically sound and computationally efficient schemes to describe aerosol formation, interaction, and evolution. In this work, we present and evaluate an aerosol dynamics module MONO32, which is designed for use in regional air pollution models. The module presents aerosol size distribution with four monodisperse modes and characterizes aerosol chemical composition with seven components, so 32 prognostic equations are needed to describe particle mass and number concentration. MONO32 accounts for nucleation, condensation, and coagulation processes. Two different time integration schemes, a FORTRAN NAG-library routine and a two-step scheme, were tested. Both integration methods showed the same accuracy in calculating particle number and size evolution, while the two-step scheme was computationally much more efficient. A mode-merging method was implemented in MONO32 to account for the transfer of aerosol mass and number to a larger mode as particles grow by condensation and coagulation. MONO32 compared reasonably well with the sectional model AEROFOR2 with 54 sections; for example, difference in the total number concentration after 24-hour simulation was less than 15–25%. MONO32 was also verified against measurements available from the Biogenic Aerosol Formation in the Boreal Forest 3 (BIOFOR3) campaign. Two typical nucleation episodes were chosen for testing, and by using the nucleation rate and the condensable vapor source rate calculated from the measurements, MONO32 was able to predict the evolution of the total number concentration and the growth rate of the nucleation mode quantitatively in good agreement with the observations. Deviations in the resulting nucleation mode diameter were 11–16%. However, the results appeared to be quite sensitive to the hygroscopic properties of the condensable organic vapor. MONO32 showed an acceptable accuracy for long-range transport modeling in describing aerosol dynamics while being computationally efficient. Therefore it is recommended for implementation and further testing in the regional three-dimensional Eulerian model.