AERMOD vs CALPUFF for Air Quality in Canada 

When is AERMOD enough, and when does a project justify CALPUFF? When people ask about air quality in Canada, one of the first technical questions can be surprisingly difficult:

Should this project be modelled with AERMOD or CALPUFF? There is no universal answer.

AERMOD and CALPUFF are both powerful air-quality dispersion models, but they are designed around different ways of representing the atmosphere. The best choice depends on the physical setting, the emissions, the distance being assessed, the regulatory objective, the available data and sometimes the history of the project.

At Calvin Consulting, AERMOD is normally our default refined model. We use CALPUFF when the project gives us a good reason to use it, such as:

• complex terrain,
• regional transport,
• deposition,
• secondary pollutants,
• a regulatory or client requirement or
• an established historical modelling approach.

Instead of choosing the most complicated model, the goal is to choose the simplest model that adequately represents the problem.

The short answer

For many industrial facilities, particularly in relatively isolated parts of Alberta, AERMOD is all that is needed.

AERMOD is a steady-state plume model that works well for many local and near-field assessments. It can handle multiple sources, buildings and downwash, complex terrain to a degree, variable emissions, deposition and a wide range of industrial source types.

CALPUFF takes a different approach. It is a non-steady-state puff model that can represent time- and space-varying meteorological conditions, curved plume trajectories, complex terrain, regional transport, chemical transformation and deposition.

So a useful first comparison is:

That table is useful, but it does not make the decision for you. The interesting part is why one model is appropriate for one project and unnecessary for another.

There is no magic distance where AERMOD becomes CALPUFF. 

A common explanation of air quality in Canada is that AERMOD is for short distances and CALPUFF is for long distances.

That's a useful introduction, but it is too simple to be a reliable modelling rule. Distance matters. It is not the only thing that matters.

A 20-km assessment through a complicated mountain valley can be a more difficult modelling problem than a longer assessment in relatively uncomplicated terrain.

The modeller needs to consider:

• transport distance;
• terrain;
• wind direction and speed;
• spatial changes in meteorology;
• source height and buoyancy;
• recirculation;
• deposition;
• chemical transformation;
• sensitive receptors; and
• the actual objective of the assessment.

British Columbia's modelling guidance illustrates this particularly well. It distinguishes between situations where a straight-line model such as AERMOD can adequately answer a near-source maximum-concentration question and situations where CALPUFF becomes more useful because the assessment needs to represent curved or spatially varying flow through complex terrain.

So instead of asking: How many kilometres is the project?, ask:

What atmospheric behaviour do we need the model to represent?

Why AERMOD is often the right starting point

AERMOD is the current EPA regulatory workhorse for refined dispersion modelling.

Assessment of air quality in Canada (Alberta)

It is a steady-state plume model that can represent both surface and elevated sources and both relatively simple and complex terrain. EPA currently maintains AERMOD as its preferred/recommended model and continues to update the model and documentation.

In Alberta, AERMOD-PRIME is one of the province's recommended refined models. Alberta describes it as a multi-source steady-state plume model.

That matters because air quality in Canada is often assessed using well-established regulatory modelling systems rather than by automatically selecting the most sophisticated tool available.

• AERMOD also has a mature supporting system:
• AERMET prepares meteorological data.
• AERMAP processes terrain and receptor elevations.
• BPIPPRM prepares building-related information for the PRIME downwash treatment.

AERMOD can also handle a range of source types and operating conditions, including constant and time-varying emissions. For a large proportion of conventional industrial facilities, that is a lot of capability.

When we normally use AERMOD

At Calvin, AERMOD is usually the starting point when the project has characteristics such as:

Relatively uncomplicated terrain - Many rural Alberta facilities fall into this category. Rolling terrain does not automatically make a problem “complex terrain.”

Local or near-field assessment - When the question concerns concentrations in the immediate area surrounding the facility, a steady-state model may adequately represent the problem.

Conventional industrial sources - Stacks, vents, process sources and other conventional source configurations are well suited to AERMOD.

No compelling need for regional transport - If the important impacts are contained within the local modelling area, there may be little benefit in introducing a regional puff model.

Appropriate regulatory framework - When the applicable provincial guidance accepts AERMOD and the physical problem does not demand something more specialized, AERMOD can provide an efficient and defensible approach.

This is one reason our default is not CALPUFF because it's more advanced. It is:

AERMOD unless there is a reason not to use it.

When do we choose CALPUFF?

At Calvin Consulting, we generally look for a specific reason before adding the additional complexity of CALPUFF. Those reasons include:

1. It was used historically: This is more important than it might first appear.

If an existing assessment program has historically used CALPUFF, maintaining consistency can sometimes be more useful than changing models simply because a newer option exists.

Historical results may also form part of a longer-term comparison. Changing models can make it more difficult to distinguish a real change in environmental impact from a change caused by the modelling method.

2. A regulator or client recommends or requires it: Sometimes the answer is simply in the project instructions.

A regulator may specify a modelling approach or a client may prefer continuity with previous work or another established modelling program. That does not eliminate the need for professional review, but it can determine the modelling pathway.

Models for forecasting air quality in rugged areas

3. The facility is close to significant complex terrain: This is one of the strongest practical reasons to consider CALPUFF.

At Calvin, we pay particular attention to projects within roughly 10 km of mountains or significant upslope terrain, particularly when the terrain is more complicated than the relatively modest rolling terrain typical of many Alberta sites.

The exact distance is not a regulatory cutoff. It is a practical screening thought: Could the terrain change the airflow enough that the straight-line assumptions of a steady-state plume model become a material limitation?

Foothills, mountain fronts, valleys and abrupt changes in elevation can produce flow patterns that deserve closer examination.

4. There are many significant emission sources: A large industrial region may create a more complicated modelling problem.

Multiple facilities can contribute to cumulative concentrations and the regional pattern may matter as much as the impact from one source. However, a large number of sources alone is not enough to justify CALPUFF.

AERMOD is perfectly capable of modelling multiple sources. The question is whether the number and distribution of sources combine with the meteorology, terrain or assessment objective to justify a more complex modelling system.

5. There is a known reason for long-range modelling: A project may require a broader assessment because meaningful impacts could occur well beyond the near field.

For Calvin, a distance somewhere around 30 km or more may be a useful indicator that the modeller should at least examine whether CALPUFF is warranted, but again, it is not a hard rule. The important question is whether transport over that distance is actually relevant to the regulatory or environmental question.

CALPUFF is specifically designed to represent non-steady-state transport over regional and longer distances and can account for time- and space-varying meteorology. EPA currently lists it as an alternative model capable of applications ranging from tens to hundreds of kilometres.

6. The assessment requires deposition or secondary pollutants: This is another major reason to look beyond a conventional AERMOD concentration assessment.

AERMOD can calculate deposition of applicable primary pollutants. But some assessments require additional treatment of:

• chemical transformation;
• secondary pollutants;
• regional transport;
• wet deposition;
• dry deposition; or
• acid deposition.

Those requirements can point toward CALPUFF or another regional modelling system. For example, an acid-deposition assessment is fundamentally different from simply predicting the ground-level concentration of SO₂ or NOx. The modeller may need to account for chemical transformation and the eventual deposition of secondary species.

This is a good example of where air quality in Canada can become a regional environmental question rather than simply a local stack-impact question.

7. The client prefers the additional capability and the budget supports it: There are occasions when a client deliberately wants a more sophisticated assessment.

That may be reasonable when the project has broader environmental objectives, when there is value in maintaining consistency with another modelling program or when the additional analysis will answer a question the client (or regulator) considers important.

But a larger budget should not, by itself, make CALPUFF scientifically necessary. More expensive modelling is not automatically better modelling.

8. Terrain and land-use details need a more sophisticated treatment: Sometimes the reason is not simply mountains. The combination of:

• terrain;
• land use;
• coastline or water bodies;
• changing surface characteristics;
• wind fields; and
• source locations

may produce a modelling problem where a spatially varying meteorological field is valuable.

CALMET, the meteorological component of CALPUFF, can develop three-dimensional meteorological fields and account for phenomena such as terrain-driven flow, valley flows, terrain blocking and lake or sea-breeze circulations.

That additional capability comes at a cost: more data, more setup and more modelling decisions.

A note about CALPUFF's EPA status

There is an important piece of history behind the AERMOD-versus-CALPUFF debate. CALPUFF was once the EPA's preferred model for long-range transport applications. That changed in 2017.

Transport of pollutants over long distances in flatlands

EPA now lists CALPUFF as an alternative model, rather than a preferred/recommended model. EPA explains that the 2017 revision established a screening approach for long-range transport and removed CALPUFF from the preferred-model list. Use of an alternative model in U.S. regulatory applications can require case-specific justification and agreement with the reviewing authority.

That does not mean CALPUFF is obsolete. It means that the old statement: CALPUFF is the EPA-approved long-range model is no longer a good description of its current U.S. regulatory status.

This history is worth understanding because it reinforces an important lesson: A model's reputation is not a substitute for checking the current guidance.

What Alberta currently recommends

Alberta's current modelling information lists: AERSCREEN for screening assessments. AERMOD-PRIME and CALPUFF for refined assessments.

Alberta describes AERMOD-PRIME as a multi-source steady-state plume model and CALPUFF as a multi-layer, multi-species, non-steady-state puff model capable of simulating time- and space-varying meteorological conditions.

Regulatory modelling in Alberta must follow the Air Quality Model Guideline 2021, and Alberta's current modelling page also identifies supplemental guidance for interpreting regulatory modelling assessments that took effect in July 2025.

That gives Alberta modellers a useful combination: Approved tools + project-specific judgement.

At Calvin, we normally start with AERMOD rather than AERSCREEN when a refined regulatory assessment is required, because AERSCREEN is a screening model rather than the refined model we would normally use for the final assessment. Alberta itself distinguishes the two levels.

What British Columbia air quality in Canada emphasizes

BC provides some of the clearest practical guidance for understanding the difference between the two modelling approaches.

AERMOD is recommended for situations such as:

• industrial source assessments;
• non-complex flow;
• buildings with or without downwash;
• deposition;
• variable emissions;
• rural or urban settings; and
• transport distances under 50 km, depending on terrain.

CALPUFF is recommended where there is:

• complex flow;
• substantial time and spatial variation in meteorology;
• complex terrain;
• coastal situations;
• local through long-range transport;
• multiple sources;
• or deposition that benefits from its additional capabilities.

Predicting if there will be air pollution in BC

One especially useful BC concept is that the assessment objective matters. AERMOD may be appropriate if you need the maximum concentration near a source. CALPUFF may be more appropriate if you need to understand the spatial distribution of impacts as a plume follows a winding valley or experiences changing winds.

That is a much more useful way to think about air quality in Canada than simply memorizing a distance cutoff.

What Saskatchewan and Manitoba tell us

Saskatchewan also recognizes both AERMOD and CALPUFF within its modelling framework and provides provincial guidance for their application.

Manitoba likewise uses refined modelling when site complexity, emissions, operating conditions or screening results indicate that more detailed assessment is needed.

The details differ between provinces, which is one reason a Canadian modeller should check the applicable provincial guidance instead of relying on a universal AERMOD-versus-CALPUFF rule. For Western Canadian projects, the broad lesson is consistent:

Start with the provincial framework, then determine whether the project's physical characteristics warrant something beyond the normal approach.

Don't choose CALPUFF just because it is more sophisticated

This is perhaps the most important practical lesson on the page. CALPUFF can do things AERMOD is not designed to do.

That does not mean CALPUFF automatically produces a better answer. A more complicated modelling system requires:

• more data;
• more preprocessing;
• more meteorological decisions;
• more chemical options;
• more parameters;
• more quality assurance; and
• more interpretation.

A model with more knobs gives the modeller more ways to make a poor choice. In other words: More sophisticated does not mean more correct.

The additional complexity should earn its way into the assessment by answering a question that matters.

Don't choose AERMOD just because it is easier either

The reverse mistake is just as important. AERMOD should not be selected simply because the modeller knows it well. If a project has:

• strong terrain effects;
• highly variable winds;
• important long-range transport;
• regional deposition;
• significant chemical transformation; or
• another feature that materially affects the assessment,

then convenience is not a sufficient reason to stay with AERMOD. A good modeller should be comfortable saying:

AERMOD is enough here and just as importantly:

AERMOD isn't enough here.

Available data can decide the question

This is often overlooked. Suppose CALPUFF appears attractive because the terrain is complicated.

• Do we have the meteorological information required to construct a defensible three-dimensional wind field?
• Do we have appropriate terrain and land-use data?
• Do we have the emissions information needed to represent all important sources?
• Do we have the information needed for any chemical or deposition calculations?

The answer matters. Do not select a model based only on what the software can do. Select it based on what the software can do with the information you actually have.

That principle applies to almost every air quality in Canada modelling project.

A typical rural Alberta example

Consider an isolated industrial facility in Alberta. The site has:

• gently rolling terrain;
• conventional stacks;
• no significant nearby mountains;
• no compelling regional chemistry question;
• impacts concentrated relatively close to the facility; and
• adequate meteorological data for AERMOD.

There is little reason to introduce CALPUFF simply because it is available. AERMOD is likely the better tool.

Now change the scenario. The facility is near the foothills, several major sources are distributed around the region, the assessment extends over a much larger area and the regulatory question includes regional deposition. 

Now the modelling problem has changed. CALPUFF may make sense. 

The important thing is that the model changed because the problem changed.

A second example: when CALPUFF still may not be necessary...

Suppose a facility is 15 km from moderately complex terrain. That fact alone does not settle the issue.

If the regulatory question is a local maximum concentration close to the facility and the terrain does not materially alter the flow in the area that matters, AERMOD may still provide an appropriate solution.

Conversely, a shorter-distance project in a narrow valley may present a stronger case for CALPUFF. This is why experienced model selection is more valuable than memorizing rules.

AERMOD of CALPUFF for air quality in Canada.

What about flaring?

Flaring provides an interesting Alberta-specific example. For many flare assessments, AERMOD remains particularly practical because the Alberta Energy Regulator's AERflare tool is closely integrated with the AERMOD modelling approach.

AERflare has been maintained and updated as part of Alberta's non-routine flare modelling framework and provides a bridge from flare characterization into AERMOD. Alberta's non-routine flaring guidance also identifies AERMOD and CALPUFF as possible refined modelling systems and describes AERflare and ABflare as tools for flare assessments.

At Calvin, that established workflow is one reason we generally prefer AERMOD for appropriate flare applications unless there is a specific reason to use another approach.

That can mean less unnecessary setup while retaining a modelling method that fits the regulatory framework.

A note about 'the best model'

Search for air quality in Canada and it is easy to find language suggesting one model is the best. That isn't how experienced modelling should work.

The better question is:

• Best for what?
• Best for a local maximum?
• Best for complex terrain?
• Best for regional transport?
• Best for deposition?
• Best for a transient flare?
• Best for a project with five years of meteorology and a straightforward rural setting?

Different answers can be correct for different questions.Common model-selection mistakes

It's over 50 km, so it must be CALPUFF. Not necessarily.

Distance matters, but terrain, meteorology and assessment objective matter too.

It's in the mountains, so AERMOD won't work. Also not necessarily.

AERMOD can treat some complex-terrain situations. The question is whether it adequately represents the particular problem.

CALPUFF is more sophisticated, so it must be more accurate. No.

Its additional capabilities are useful only when they are relevant and supported by appropriate data.

We've always used AERMOD. Historical practice is useful—but not a substitute for examining the current project.

The regulator accepts CALPUFF, so we'll use it. Acceptance does not automatically make it necessary.

We already have AERMOD inputs, so let's not complicate things. Convenience is not a scientific basis for model selection.

More modelling must mean a better assessment. Not necessarily.

A well-designed, appropriately scoped AERMOD assessment can be more defensible than an elaborate CALPUFF study built on poorly supported assumptions.

The Calvin approach

At Calvin Consulting, we do not begin with: Which model do we sell?

We begin with: What is this project asking us to predict? For many facilities, the answer leads to AERMOD.

For others, the characteristics of the terrain, meteorology, transport distance, emissions, chemistry or regulatory objective lead us toward CALPUFF.

Sometimes the right answer is another accepted modelling approach entirely. That judgement is part of the job.

We also recognize that the model itself is only one component of a defensible assessment. Source parameters, terrain, meteorology, receptor selection, building representation and background information can all matter at least as much as the choice between AERMOD and CALPUFF.

The Takeaway

AERMOD and CALPUFF are not competing answers to a multiple-choice question. They are different tools for different modelling problems.

Choose AERMOD when it adequately represents the source, atmosphere, terrain, scale and regulatory question.

Consider CALPUFF when the project's complexity makes its additional capabilities genuinely useful.

And before selecting either one, ask: What does the assessment actually need to represent?

That question will usually tell you more than a distance cutoff or a list of software features.

Need help choosing the right model?

The most difficult part of an air-quality modelling project is often not running the model. It is deciding what should be modelled, how it should be represented and which assumptions can be defended.

Calvin Consulting Group has more than 30 years of experience with industrial dispersion modelling across Western Canada. AERMOD is our normal starting point, while CALPUFF and other approaches are considered when the project presents a genuine reason to use them.

That means a client does not need to become a dispersion modeller simply to figure out which software to buy, which settings to use or whether a more complicated assessment is necessary.

A few hours of experienced model selection can save weeks of modelling the wrong problem.

Whether the project involves a conventional rural Alberta facility, complex terrain, regional transport, cumulative effects, deposition or unusual emissions, we can help establish an appropriate modelling approach before the detailed work begins.

Contact Calvin Consulting Group to discuss your project and determine which air-quality modelling approach makes sense.

Clean air is our Passion...Regulatory Compliance is our Business.

The model is only part of the assessment

The software is important. But so are:

The emissions - Are the emission rates representative?

The meteorology - Does it represent the site and the time period?

The terrain - Is the landscape represented correctly?

The buildings - Does BPIP-PRIME or the selected downwash treatment describe the facility properly?

The receptors - Are we looking where people and environmental receptors actually are?

The background - Are existing sources and ambient concentrations represented appropriately?

The interpretation - Does the result make physical and regulatory sense?

A technically sophisticated model cannot rescue a poor representation of the underlying facility. That is why professional dispersion modelling is more than choosing software and pressing run.

AERMOD or CALPUFF? A practical decision tree

Here is the way I would reduce the choice to a series of questions. Start with the regulatory question

What are we actually being asked to predict?

A local maximum concentration?

A spatial concentration pattern?

Regional transport?

Deposition?

Secondary pollutants?

Cumulative effects?

↓

Is the problem primarily local and relatively uncomplicated?

Yes → AERMOD is a strong starting point.

No → continue.

↓

Could terrain materially alter atmospheric flow?

Mountains?

Foothills?

Valleys?

Strong upslope effects?

Coastal circulation?

Yes → CALPUFF deserves serious consideration.

↓

Does the assessment depend on time- and space-varying meteorology?

Yes → CALPUFF becomes more attractive.

↓

Is regional transport or deposition important?

Yes → consider CALPUFF or another appropriate regional model.

↓

Do we actually have the data to support the more complicated approach?

This is critical.

No → reconsider whether the more complex model can be justified.

Yes → proceed with the model that best represents the problem.

The final question is always:

Can we defend the model scientifically and regulatorily?

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, and the author remains responsible for the content.

What an experienced modeller checks before choosing

At Calvin, our model-selection process starts with the project rather than the software.

We look at:

Terrain - How much relief is there? Are there mountains, steep slopes, valleys or other features that could affect flow?

Source characteristics - Are the sources tall, buoyant, directional, intermittent or unusually configured?

Transport distance - How far can the assessment reasonably extend?

Meteorology - Can the atmosphere be represented adequately with a steady-state approach or do we need spatially varying fields?

Receptors - Are there communities, sensitive ecosystems or other locations where the spatial pattern matters?

Chemistry - Are primary pollutants enough or do secondary pollutants need to be represented?

Deposition: - Is ground-level concentration the endpoint or do we need wet or dry deposition?

Existing modelling - Has the project already been assessed using a particular model?

Regulatory expectations - What does the applicable provincial guidance require or recommend?

Data quality - Can the chosen model actually be supported with representative data?

These questions often make the model choice much clearer.