AERMOD is the U.S. EPA preferred air dispersion model. The development of AERMOD started in 1991, based on the indications of AERMIC (American Meteorological Society / Environmental Protection Agency Regulatory Model Improvement Committee) that outlined a new basis for steady state air quality models to be used for regulatory purposes. Starting from December 2007 AERMOD has replaced ISC3 among the models recommended by U.S. EPA for modelling the impact of ground level and elevated industrial sources on flat or moderately complex terrain.
AERMOD can simultaneously simulate many sources with different shapes, at ground or elevated, buoyant or non buoyant, emitting one or more pollutants. AERMOD is capable to account for the non homogeneous vertical structure of the boundary layer (also through the use of a vertical profile of meteorological variables). Vertical mixing is limited in case of stable conditions. The dispersion for unstable conditions is non-Gaussian, so to correctly describe the high concentrations of pollutants that can be observed close to stacks under convective conditions.
AERMOD is made of different modules:
AERMOD requires two sets of meteorological data, one at surface and the other referring to a vertical profile, both with hourly time resolution. The required variables at surface are: sensible heat flux, friction velocity, convective velocity, vertical temperature gradient in the first 500 m above the planetary boundary layer, the extent of the convective boundary layer, the extent of the mechanical boundary layer, the Monin Obukhov length, the surface roughness, the Bowen ratio, the albedo, the wind speed, the wind direction, the anemometer height, the temperature, the thermometer height. Variables included in the vertical profile are, for each elevation above ground, the elevation itself, the wind speed, the wind direction, the temperature, the standard deviation of wind direction and the standard deviation of vertical wind speed.
It is very important to have a complete picture about wind direction and wind speed when doing an atmospheric dispersion simulation. The WindRose PRO3 software allows to create wind roses directly from the AERMOD input file containing meteorological data at surface.
A screening version of AERMOD, called AERSCREEN, has been developed by the US-EPA in order to estimate the worst possible ground level concentrations (GLCs) from a single source. In order to facilitate the use of AERSCREEN Enviroware has developed a Graphical User Interface called AerscreenEZ.
AERMOD includes several improvements compared to standard Gaussian models:
Standard Gaussian models are based on six discrete stability classes (Pasquill Gifford classes) to which correspond dispersion parameters that are based on observations from ground level releases. On the contrary, AERMOD uses vertical continuous profiles of horizontal and vertical turbulence that are either based on measurements or computed based on similarity theory.
AERMOD describes the non Gaussian vertical dispersion under convective conditions, that are characterised by the presence of updraft and downdraft motions with different probability of occurrence and different intensity. Under convective conditions the plume is made of three components: a direct plume that is brought to the ground by a downdraft, an indirect plume that is captured by an updraft up to the superior lid and is then possibly brought downwards by a downdraft, and a third plume penetrating the mixing lid and dispersing more slowly in the stable layer above and possibly re-enter in the mixing layer and reach the ground.
Under stable conditions AERMOD describes the horizontal and vertical dispersion in the same way standard Gaussian models as ISC3 do. However, while models as ISC3 assume an infinite boundary layer, AERMOD accounts for the possible reflections by a superior lid.
Standard Gaussian models generally use the Briggs equations for calculating the effective height of the release due to the buoyancy of the plume with the wind speed and temperature gradient values at the stack height. Instead, under stable atmospheric conditions AERMOD uses the values at stack height at half distance from the final height due to buoyancy, while under convective conditions it superimposes the random displacements due to the random fluctuations of the convective velocities.
Sources can be treated as rural or urban independently.
AERMOD has a terrain processor (AERMAP) that prepares the data for their use within the model by advanced algorithms that discriminate the streamline division based on a critical height.
Sources: EPA-454/R-03-004 (2004) AERMOD: Description of model formulation.