Abstract. The aerosol dynamics module MADE has been coupled to the general
circulation model ECHAM4 to simulate the chemical composition,
number concentration, and size distribution of the global submicrometer
aerosol. The present publication describes the new model system
ECHAM4/MADE and presents model results in comparison with
observations. The new model is able to simulate the full life cycle of
particulate matter and various gaseous particle precursors including
emissions of primary particles and trace gases, advection, convection,
diffusion, coagulation, condensation, nucleation of sulfuric acid vapor,
aerosol chemistry, cloud processing, and size-dependent dry and wet
deposition. Aerosol components considered are sulfate (SO4),
ammonium (NH4), nitrate (NO3), black carbon (BC),
particulate organic matter (POM), sea salt, mineral dust, and aerosol
The model is numerically efficient enough to allow long
term simulations, which is an essential requirement for application in
general circulation models.
Since the current study is focusing on the submicrometer aerosol, a
coarse mode is not being simulated. The model is run in a
passive mode, i.e. no feedbacks between the MADE aerosols and clouds
or radiation are considered yet. This allows the investigation of the
effect of aerosol dynamics, not interfered by feedbacks of the altered
aerosols on clouds, radiation, and on the model dynamics.
In order to evaluate the results obtained with this new model system,
calculated mass concentrations, particle number concentrations, and
size distributions are compared to observations.
The intercomparison shows, that ECHAM4/MADE is able to reproduce the major
features of the geographical patterns, seasonal cycle, and vertical
distributions of the basic aerosol parameters. In particular, the
model performs well under polluted continental conditions in the
northern hemispheric lower and middle troposphere. However, in
comparatively clean remote areas, e.g. in the upper troposphere or in
the southern hemispheric marine boundary layer, the current model
version tends to underestimate particle number concentrations.