The AERMOD Model Pushes Environmental Science into a New Era.

The AERMOD model is a modern tool for predicting air pollution spread. A group of specialists created it and went through a bit of a process to make sure it works. Before the Environmental Protection Agency (EPA) approved it, it was tested, reviewed, and improved.

AERMOD uses weather data and terrain info to predict the distribution of pollution. Using modern knowledge about how air moves and interacts with land, it has replaced older models. Experts can make better decisions that impact air quality than they could before.

This model helps us understand how pollution spreads in cities as well as rural and marine areas. As updated science comes out, it keeps getting better, and so does our environment. With this page on AERMOD, you can see how modern technology is changing how we manage air quality and tackle environmental problems.

AERMOD - What's under the hood?

We start with a history of the AERMOD model.  Understanding the planetary boundary layer (i.e., the PBL, the air closest to the ground) is crucial for improving regulatory dispersion models. This turbulent layer of air affects pollution dispersion most greatly. The American Meteorological Society (AMS) and EPA formed the AMS/EPA Regulatory Model Improvement Committee (AERMIC) to improve existing models.

Dynamics of Air DispersionBeautiful Outdoor Air Quality

What are models?  Computer programs designed to indicate how pollutants will move through the air.

AERMIC incorporated PBL concepts into regulatory models in the 1970s and 1980s. Through numerical modeling, field observations, and laboratory simulations, this period saw breakthroughs in understanding convective and stable boundary layers.

Using these insights along with dispersion techniques and models, the committee aimed to create a state-of-the-art regulatory model. This initiative addressed other critical factors like terrain interactions and urban dispersion, which also needed to be addressed.

The Previous Air Quality Dispersion Model

AERMIC replaced the EPA's Industrial Source Complex Model (ISC3) for near-field air pollution impact assessments. AERMIC developed AERMOD in response to ISC3's outdated algorithms. Using state-of-the-art modeling techniques to represent atmospheric dispersion, this new model would keep ISC3's user-friendly input/output structure.

AERMOD has three main components: AERMOD itself, the dispersion model; AERMET, the meteorological preprocessor; and AERMAP, the terrain preprocessor. With meteorological data, AERMET characterizes the planetary boundary layer, while AERMAP calculates terrain influence.

AERMOD is designed to be accurate, user-friendly, and adaptable. As compared to ISC3, AERMOD offers the following improvements:

  • Pollution dispersion in the atmospheric boundary layer, the part of the air closest to the ground influenced by turbulence.
  • Plume rise - When emissions rise vertically due to buoyancy and momentum
  • Pollutant dispersion and flow influenced by terrain features
  • Pollutant dispersion and concentration caused by turbulent wakes created by buildings

These are based on the latest scientific understanding of atmospheric processes.

Creation of the AERMOD Model

The process of AERMOD's development involved model formulation, development evaluation, internal peer review, beta testing, revised model formulation, performance evaluation, sensitivity testing, external peer review, and regulatory consideration.

AERMOD input informationSeamless integration

The developers used field measurements to identify areas for improvement. The model was also tested against independent databases and other regulatory models. After public comments and further revisions, AERMOD was proposed as the EPA's preferred regulatory model. Here's a brief description of the AERMOD model.

In addition to urban and rural settings, complex terrains, and limited overwater scenarios, AERMOD provides air dispersion predictions across diverse environments. It handles various sources, making transitions smooth without sudden concentration changes. This new AERMOD model has been enhanced for specific emission scenarios:

  • Identifying and defining the physical and chemical properties of emission sources for accurate dispersion modeling
  • Modeling pollution dispersion over water bodies using offshore meteorology
  • BUOYLINE and RLINE - Modeling emissions from buoyant line sources and roads

AERMOD uses a steady-state plume model, which includes Gaussian distributions in stable conditions and bi-Gaussian probability density functions (PDFs) in convective conditions.

This simplified approach eliminates the need to categorize terrain types. It can also characterize the PBL from readily available data. With the AERMOD model, the AERMET meteorological preprocessor, and the AERMAP terrain preprocessor, the integrated AERMOD modelling system streamlines the pollutant dispersion prediction process.

AERMET

AERMOD relies on AERMET to convert raw weather data into information AERMOD can use to predict pollution spread. In AERMET, wind, temperature, and other weather conditions are measured and important factors about the atmosphere are calculated, like how stable or turbulent it is.

Wind profiling that worksMeteorology and Pollution Dispersion

When you think about how rough the ground is and how much sunlight it reflects, you can see how it affects the air above. These details help AERMOD create accurate pictures of air movement and pollution dispersal.

With more granular information, AERMET becomes more accurate. This description applies specifically to land-based weather; ocean weather requires a unique set of calculations.  Search the EPA Support Center for Regulatory Air Models (SCRAM) website for COARE.

What's special about the meteorology of the PBL? Understanding the PBL, the lowest part of the atmosphere and the one most directly influenced by the Earth, is essential to modeling air pollution dispersion. Surface sensible heat flux (H) tells us how much heat is transferred from the ground to the air, in units of energy per area per unit of time. AERMET calculates this.

Heat flux = sensible heat flux + latent heat flux + soil heat flux

i.e. Rn = H+ λE+G

The sun's energy (net radiation, Rn) gets distributed differently: some heats the air (H), some evaporates water (latent heat flux, λE), and some goes into the ground (soil heat flux, G). AERMET estimates these fluxes, especially H, to determine how the PBL behaves. In the absence of direct measurements, AERMET can estimate Rn. It even determines the time when the sun warms the atmosphere, leading to a convective boundary layer (CBL) and when it cools down at night, and thus a stable boundary layer (SBL).

AERMET determines the moment of transition from the daytime CBL to the nighttime SBL by calculating the critical solar elevation angle (φcrit). At this angle, heat flux reverses, signaling a shift. AERMET calculates φcrit based on temperature, cloud cover, and solar radiation. It's usually an angle of 13° on clear days and 23° on cloudy days, or you can use equivalent cloud cover data if available. 

Why is CBL in the atmosphere important?

It's important to understand turbulence and how critical parameters in the CBL are calculated. AERMET refines the estimates iteratively (i.e., a small step at a time), starting with neutral conditions and refining them until they agree.

Transformations in the atmosphereMastering Solar fluxes and Heat fluxes

I'll now introduce the Monin-Obukhov length, represented by L. It helps predict how pollutants disperse under different atmospheric stability conditions since buoyancy dominates at a height equal to L.

As the wind moves across the ground, friction occurs, and theorists compare with L to tell us how stable or unstable the air is. AERMET uses mathematical formulas and iterative methods to get the best fit. They find a weighted average that represents the speed of large turbulent eddies in the CBL. When the atmosphere is convective during the day, AERMOD uses these calculations to predict how pollutants will disperse.

And at night?  SBL in the Atmosphere...

The semi-empirical approach used in AERMET avoids the complexities of nocturnal energy balance calculations during stable atmospheric conditions, typically at night. 

AERMET calculates wind profiles based on observations and mathematical derivations, using a relationship between temperature scale and wind profiles. A cloud cover estimate (θ*) is made directly if cloud cover data is available; otherwise, it's calculated using temperature measurements at two levels and wind speed at one.

An alternative approach based on the work of researchers, Qian and Venkatram (2011) has been added since the 2017 version of the AERMOD model and its update to AERMET.  It applies especially for urban areas. Based on the friction velocity (u*) and θ*, AERMET calculates the sensitive heat flux (H) and then L. In order to prevent unrealistic estimates of heat loss, AERMET limits the heat flux and adjusts u* accordingly. AERMOD relies on these calculations to simulate pollutant dispersion accurately.  

Friction velocity measures how much shear stress the wind has near the ground.  Since it helps determine how pollutants mix and disperse, it plays a vital role in modern air quality modeling.

What is Mixing Height?

For pollutant dispersion modeling, zi represents the vertical extent of turbulent mixing in the atmosphere. Convective (zic) and mechanical (zim) turbulence determine mixing height in CBL. Since only mechanical turbulence matters in the SBL, zi equals zim at night. 

Meteorological FlexibilityInsights into heights and dispersion

AERMET calculates zic using an energy balance model that accounts for early morning temperatures. "zim" calculates an equilibrium height, then smooths it over time. In AERMOD's calculations, zi is used as a reflecting/penetrating surface and this smoothing process ensures a continuous transition and prevents discontinuities.

Buildings, mountains, and other structures disrupt the natural flow of air near the ground, causing irregular air motion and mechanical turbulence while convective turbulence forms when the Earth's surface is unevenly heated, causing warm air to rise and cool air to sink.

How does AERMET account for stable conditions and low winds?

In calm, stable weather, standard modelling calculations might overestimate pollutant concentrations. A special adjustment changes how AERMET calculates u* under these conditions.

To use it, a modeller can set ADJ_U* in AERMET's settings. AERMET uses different formulas depending on whether you select a Bulk Richardson Number (which describes temperature differences). If you don't use Bulk Richardson, the program chooses a formula from either Qian and Venkatram (2011) or Luhar and Rayner (2009) formulas described in the model documentation. This flexibility helps AERMET make more accurate predictions when the wind is light and the atmosphere is stable.

  • It's best to use the Bulk Richardson Number when temperature gradient data is available.
  • With surface characteristics and wind data, Qian/Venkatram estimates u* for you
  • If you have reliable temperature gradient data, the Luhar/Rayner formulation gives even better results.

The biggest effect is a reduction in predicted concentrations in stable, low-wind conditions. These algorithms can improve the correlation between AERMOD predictions and observed concentrations in stable environments, however they are highly dependent on the quality and representativeness of the input meteorological data.

PBL and vertical structure

AERMOD uses its internal meteorological interface to build a detailed picture of the atmosphere. This interface calculates wind direction, wind speed, temperature, temperature gradients, and turbulence vertical profiles using AERMET's boundary layer parameters.

Dynamics of diffusionWhat does Turbulence Mean?

The interface compares the heights where AERMOD needs meteorological data with actual measurements. Without data, established scientific formulas (similarity relationships) are used to estimate values. For better accuracy, measurements above and below a point are interpolated, combining the measured data with the calculated profile. The AERMOD model acts as though it understands the vertical layers of the atmosphere even without direct measurements.

What Profiles?

AERMOD's meteorological interface lets you build detailed vertical profiles of atmospheric conditions. In order to build these profiles, AERMET provides site-specific information and mathematical relationships. AERMOD can estimate conditions even when direct measurements aren't possible.

When calculating wind speed, AERMOD adjusts for surface roughness and atmospheric stability. For intermediate heights, linear interpolation is used. Potential temperature gradients are calculated differently for stable and convective conditions to understand atmospheric stability. A combination of mechanical and convective contributions is used to model pollutant dispersion. AERMOD needs these profiles to simulate pollution transport and diffusion.

Boundary Interface

To get accurate pollutant dispersion modeling, AERMOD accounts for the fact that wind and turbulence vary with height. As AERMOD is a steady-state model, these vertical variations have to be represented with single, "effective" values. Weather conditions are averaged over the portion of the atmosphere where the plume travels.

Think of the plume as a cloud of pollution. AERMOD calculates effective wind speed, turbulence, and temperature gradients by averaging the conditions within the plume's vertical extent. As a result of this averaging, the model accurately reflects pollutant transport conditions. The averaging window is determined by the centerline and 2.15 standard deviations of vertical dispersion up and down.

The Model, AERMOD

By using sophisticated air dispersion models, the AERMOD model simulates pollution spread in the atmosphere. There's no need to categorize terrain because the model seamlessly handles both flat and complex terrain. It calculates concentrations based on hourly weather averages for understanding overall pollution distribution.

Sum of two extreme plume statesVarious conditions for pollutant dispersion

The calculations of AERMOD depend on atmospheric stability. The model assumes a Gaussian distribution for both horizontal and vertical pollutant concentrations. When the horizontal distribution stays Gaussian, a bi-Gaussian probability density function accounts for phenomena like "plume lofting." Plume lofting is when pollution rises into the atmosphere and spreads because the air above this pollution is warmer and less dense, enabling the plume to travel upwards even faster.

AERMOD enhances turbulence to reflect the unique convective-like boundary layer that forms at night near clustered sources of urban heat. To model plume interaction with hills and valleys, AERMOD uses the concept of a dividing streamline based on atmospheric stability and plume height.

The AERMOD Model's Structure

AERMOD models plumes as a combination of horizontal, terrain-impacting plumes and terrain-following plumes. On flat terrain, these states are the same. To calculate the total concentration at a specified location (called a receptor) in complex terrain, AERMOD uses a dividing streamline height (Hc).

AERMAP calculates a terrain-influenced Hc for each receiver, so the AERMOD model can handle flat and complex terrain. The weighting it uses for influences depends on Hc and the vertical concentration distribution, which is affected by plume height, receptor elevation, and atmospheric stability.

With AERMOD, you can handle direct, indirect, penetrated, injected, and stable plumes. It also accounts for vertical variations in wind speed and turbulence.

Results in CBL

AERMOD uses an advanced dispersion formulation in the CBL. It models plumes moving through updrafts and downdrafts based on probability patterns. Since vertical air movement is skewed, the vertical concentration distribution in the CBL isn't perfectly Gaussian. Because updrafts and downdrafts affect plume dispersion differently, AERMOD approximates this skewness with the bi-Gaussian distribution mentioned earlier.

Dealing with desert terrainDynamics of plumes

AERMOD accounts for buoyant releases by tracking plume trajectories based on plume rise and random vertical displacements. An "indirect" source above the mixed layer (zi) represents plume lofting, and a "penetrated" source represents material that enters a stable layer above the CBL. To satisfy boundary conditions, CBL predictions are based on these sources and virtual sources.

Results in SBL

AERMOD uses a Gaussian plume model to calculate pollutant concentrations in an SBL. Pollutants spread horizontally and vertically in an approximately bell shaped pattern. To account for limited turbulence above the mechanical mixing layer (zim), AERMOD introduces an effective mixing lid (zieff).

It acts as a flexible boundary that restricts plume material from rising further, but not completely. Its height is calculated by multiplying the plume height by 2.15 times the vertical standard-deviation. By simulating plumes' interactions with the stable layer above zim, AERMOD reduces the impact of low-level turbulence on vertical dispersion to prevent unrealistic concentration increases in its estimates.

Lateral Meandering

As a result of low-frequency eddies, AERMOD accounts for plume meander, which affects concentration predictions at greater distances. There are two concentration limits in AERMOD: the coherent plume limit, where the plume follows a defined wind direction, and the random plume limit, where it's equally likely to appear anywhere.

Interpolation is based on the idea that horizontal wind energy is distributed between mean and turbulent components. Based on the ratio of random wind energy to total wind energy, AERMOD calculates the relative contribution of each concentration limit.

Technical Topics in AERMOD Modelling

We start with Dispersion Coefficients

With AERMOD, you can calculate pollution dispersion by combining ambient turbulence and buoyancy-induced turbulence. "Effective parameters" are used to account for ambient turbulence that varies with height. AERMOD models both lateral and vertical dispersion using surface dispersion treatments and Taylor's theory.

Mastering the plume riseCharacterizing sources for modeling

AERMOD incorporates Plume Rise Model Enhancements (PRIME) when buildings influence plume behavior. Buildings create wakes that enhance plume growth and restrict plume rise in PRIME. As the plume moves away from the building, PRIME's near-wake calculations and AERMOD's far-field predictions gradually shift from PRIME to AERMOD dominance.

Plume Rise?

AERMOD calculates plume rise based on atmospheric stability. In the CBL, plume rise is determined by momentum and buoyancy, using equations that take wind speed and stack characteristics into account. To model plume lofting, the reflected plume height is adjusted for the indirect source. The plume height of a penetrated source is calculated in a stratified environment.

An iterative approach is used by AERMOD to calculate plume rise in the SBL, which accounts for plume buoyancy decreasing as it rises. By repeatedly averaging wind speed and stability parameters, the plume rise converges. Additionally, the AERMOD model incorporates limits on plume rise in near-neutral and near-calm conditions.

Characterization of Emission Sources

The AERMOD model can handle point, volume, area, roadway, buoyant line, and irregularly shaped emission sources. It handles volume sources differently than ISC3, which handles point and volume sources the same way. Aside from squares and rectangles, area sources can be circular or polygonal, and half-life decay can be calculated.

AERMOD integrates algorithms from the Buoyant Line and Point (BLP) model, requiring inputs on building dimensions and source dimensions. AERMOD's RLINE source type is an adaptation of the R-LINE model, which has been updated to work with other AERMOD types.

PVMRM and Nitrogen Dioxide (NO2)

In AERMOD, the Plume Volume Molar Ratio Method (PVMRM) converts NOx to NO2 by including the amount of ozone in the plume volume. It uses relative dispersion coefficients, which reflect the instantaneous spread of the plume, especially in unstable conditions. These coefficients combine buoyancy and ambient turbulence with volume and area adjustments.

Need PVMRM?Dispersion and Conversion with Precision

To address overpredictions in stable conditions, PVMRM now uses total dispersion coefficients. For unstable conditions, 2.58 sigmas (standard deviations for defining a width) were used, and for stable conditions, 1.282 sigmas.

Terrain effects and plume penetration are also taken into account. In addition, AERMOD applies a minimum ozone concentration based on the Monin-Obukhov length, which can be overridden for surface releases and shorter stacks.

GRSM

The Generic Reaction Set Method (GRSM) simulates the complex chemical reactions between nitrogen oxides (NOx) and ozone (O3). Based on the Atmospheric Dispersion Model Method (ADMSM), the GRSM uses a semi-empirical photochemical model, Generic Reaction Set (GRS), to account for NO, NO2, and O3 dynamic equilibrium. Based on solar radiation and plume travel time, this method calculates reaction rates. From these partitioned concentrations, ensemble plume concentrations are calculated, reflecting chemical transformations.

Urban Effects

Due to the urban heat island effect, AERMOD accounts for the unique characteristics of cities. We boost turbulence in the stable boundary layer by using urban-rural temperature differences and population sizes. The model calculates urban mixing height and convective velocity scale based on friction velocity and Monin-Obukhov length.

Most rural formulations are used during the day, but adjustments are made in the early morning when rural convective mixing heights are lower than urban mixing heights. Adjustments for urban boundary layers have been added for RLINE, RLINEXT, and BUOYLINE sources.

Symbols used within Model

A comprehensive list of variables and constants is provided in the AERMOD air dispersion model documentation. It's easier to understand model equations and processes if you understand the symbols. Dispersion types, crucial for understanding pollution spread, are represented by sigma σ, while friction velocity, key to atmospheric turbulence, is represented by u*.

Alphabet and numeric innovationsClarification of regulations and data

Meteorological variables like wind speed u and temperature T to model-specific terms like plume rise Δh and concentration contributions C. There are also dimensionless parameters and constants in the model. By understanding these and the other symbols, you'll understand how AERMOD simulates pollution dispersion.

Data requirements

AERMOD uses hourly meteorological data inputs like National Weather Service data, morning soundings, and site-specific measurements. The wind speed, wind direction, cloud cover, ambient temperature, and morning soundings are all vital. Users should use AERSURFACE if possible to define surface characteristics like albedo, Bowen ratio, and roughness length. We also need latitude, longitude, and time zone info, as well as wind speed thresholds. You can also input turbulence profiles and solar and net radiation.

Data Processing and Use

To create meteorological profiles, AERMOD measures winds, temperatures, and turbulence. Site-specific data requires a threshold wind speed, and reference heights for wind and temperature are determined by data availability. Sounding data is used to calculate potential temperature gradients above mixing height. To prevent unrealistic results, AERMOD uses measured turbulence values with minimum limits. AERMET doesn't default to off-site data when there's no on-site data.

Transfer and model adjustments

AERMET gives AERMOD detailed weather info. The maximum height of the PBL is 4000 meters. The mechanical mixing height smoothing is initialized with specific conditions. When sounding data isn't enough, AERMET extrapolates the potential temperature gradient. AERMAP sends AERMOD receptor and terrain data. During model calculations, wind speed and turbulence limits are applied. Based on specific rules for stable and convective conditions, AERMET can substitute calculated mixing heights for missing measurements.

Meteorological Interpolation

AERMOD creates continuous meteorological profiles by interpolating between observations. As a result, the model accurately represents vertical variation in meteorological parameters. As for profile heights between observed levels, AERMOD does weighted linear interpolation, adjusting the interpolated values based on similarity functions. AERMOD extrapolates the nearest observation to maintain profile consistency beyond the range of observations.

Measured mixing heights

AERMOD uses measured mixing heights in its calculations. In stable boundary layer conditions (L>0), the measured mixing height is used as a mechanical mixing height (zie). When the measured mixing height (zic) is used to calculate and smooth the equilibrium height (zie), a convective boundary layer forms. Calculated mixing heights can be used instead of missing measured values, and substitutions are logged.

Alpha and Beta Options

AERMOD has "ALPHA" options that let users explore and test experimental features beyond the standard regulatory formulation. The AERMOD model doesn't automatically use these options, but it offers opportunities for research and advanced applications. Users can use ALPHA options to investigate different aspects of air pollutant dispersion, for example. Refer to the current AERMOD User's Guide for instructions on how to add these features.

What are options?

AERMOD has BETA and ALPHA options for exploring model enhancements at different stages. Beta options are well-established scientific advances that are fully implemented and validated, and are ripe for regulatory approval. Alternatively, the ALPHA options are in an earlier stage, ranging from concept ideas to near-mature features. Prior to going through a formal regulatory process, the ALPHA options represent potential scientific changes to AERMOD.

Innovators

ALPHA options are experimental, less mature than BETA, ranging from early concepts to near-final solutions. Regulatory approval is required before they're included in the official model. Here are the main ALPHA options:

  • You can model depressed roadways, roadside barriers, and wind profiles without displacement height with RLINEXT.
  • Improvements to downwash:
  • ORD_DWNW fixes plume width mismatch, effective wind speed, and maximum turbulence.
  • AWMADWNW focuses on effective wind speed, turbulence enhancement, velocity deficit, and entrainment constants for streamlined buildings.
  • Platform downwash algorithm based on OCD model.
  • SWPOINT: A new source type for studying building wake cavities.

Model performance can be improved by adjusting turbulence parameters (sigma-v, sigma-w), minimum wind speed, and plume meander calculations. You can adjust the weighting of the random plume contribution.

The plume meander is calculated using momentum balance instead of energy balance.

When Your Business Needs Air Quality Dispersion Modelling...

Air Quality Testers at Calvin Consulting Group Ltd.Air quality dispersion modeling by Calvin Consulting

When air quality dispersion modelling stands between you and regulatory approval, the last thing you need is complexity. You need results—fast, precise, and defensible and That’s exactly what we deliver at Calvin Consulting Group Ltd.

Our seasoned meteorologists and dispersion modellers—some of the most experienced in Canada—have spent decades perfecting the science of air quality assessments. We don’t just run models; we train regulators, industry leaders, and engineers on how to use them effectively. With over 95 years of combined expertise, our team provides the insight and accuracy required for regulatory compliance and risk mitigation.

With tools like AERMOD (and CALPUFF), we build site-specific meteorological datasets, integrate terrain and emissions data, and generate reports that are not only compliant but clear, compelling, and ready for submission. Whether you’re securing approvals, assessing emergency risks, or ensuring environmental responsibility, our assessments stand up to scrutiny and keep your project moving forward.

Your time is valuable. Your reputation matters. Trust a team that understands both. Contact Barry at Calvin Consulting Group Ltd. today and let’s simplify the path to your air quality compliance.

✉ Get Barry's help now, you'll be glad you did

...for quicker, more detailed answers. 

A certified Environmental Professional (EP) will provide you with the best assistance on your project.

New! Comments

Do you like what you see here? Please let us know in the box below.

With this air quality model, we're transforming our future.

American scientists developed the AERMOD model in seven steps, including creating it, testing it, evaluating its performance, beta testing, and getting public feedback. AERMET and AERMAP process weather and terrain data to help predict pollution movement. In replacing the ISC3 model, this design challenges old ideas and brings in a new era of environmental protection.


Do you have concerns about air pollution in your area??

Perhaps modelling air pollution will provide the answers to your question.

That is what I do on a full-time basis.  Find out if it is necessary for your project.


Have your Say...

on the StuffintheAir         facebook page

Other topics listed in these guides:

The Stuff-in-the-Air Site Map

And, 

See the newsletter chronicle. 

Thank you to my research and writing assistants, ChatGPT and WordTune, as well as Wombo and others for the images.

GPT-4, OpenAI's large-scale language generation model (and others provided by Google and Meta), helped generate this text.  As soon as draft language is generated, the author reviews, edits, and revises it to their own liking and is responsible for the content.