Development and Evaluation of the Aerosol Dynamics and Gas Phase Chemistry Model ADCHEM

P. Roldin,E. Swietlicki,G. Schurgers,A. Arneth,K. E. J. Lehtinen,M. Boy,M. Kulmala

DOI: https://doi.org/10.5194/acp-11-5867-2011

IF: 6.3

2011-01-01

Atmospheric Chemistry and Physics

Abstract:The aim of this work was to develop a model suited for detailed studies of aerosol dynamics, gas and particle phase chemistry within urban plumes, from local scale (1 × 1 km2) to regional scale. This article describes and evaluates the trajectory model for Aerosol Dynamics, gas and particle phase CHEMistry and radiative transfer (ADCHEM). The model treats both vertical and horizontal dispersion perpendicular to an air mass trajectory (2-space dimensions). The Lagrangian approach enables a more detailed representation of the aerosol dynamics, gas and particle phase chemistry and a finer spatial and temporal resolution compared to that of available regional 3D-CTMs. These features make it among others well suited for urban plume studies. The aerosol dynamics model includes Brownian coagulation, dry deposition, wet deposition, in-cloud processing, condensation, evaporation, primary particle emissions and homogeneous nucleation. The organic mass partitioning was either modeled with a 2-dimensional volatility basis set (2D-VBS) or with the traditional two-product model approach. In ADCHEM these models consider the diffusion limited and particle size dependent condensation and evaporation of 110 and 40 different organic compounds respectively. The gas phase chemistry model calculates the gas phase concentrations of 61 different species, using 130 different chemical reactions. Daily isoprene and monoterpene emissions from European forests were simulated separately with the vegetation model LPJ-GUESS, and included as input to ADCHEM. ADCHEM was used to simulate the ageing of the urban plumes from the city of Malmö in southern Sweden (280 000 inhabitants). Several sensitivity tests were performed concerning the number of size bins, size structure method, aerosol dynamic processes, vertical and horizontal mixing, coupled or uncoupled condensation and the secondary organic aerosol formation. The simulations show that the full-stationary size structure gives accurate results with little numerical diffusion when more than 50 size bins are used between 1.5 and 2500 nm, while the moving-center method is preferable when only a few size bins are selected. The particle number size distribution in the center of the urban plume from Malmö was mainly affected by dry deposition, coagulation and vertical dilution. The modeled PM2.5 mass was dominated by organic material, nitrate, sulfate and ammonium. If the condensation of HNO3 and NH3 was treated as a coupled process (pH independent) the model gave lower nitrate PM2.5 mass than if considering uncoupled condensation. Although the time of ageing from that SOA precursors are emitted until condensable products are formed is substantially different with the 2D-VBS and two product model, the models gave similar total organic mass concentrations.