CALMET and CALPUFF Modelling: Where Expert Judgement Matters

Air dispersion modelling helps industrial operators understand how emissions may affect air quality beyond their facility boundaries. Yet the reliability of a modelling assessment depends on more than running a model correctly; it takes expert judgement. Meteorological data, terrain, emissions, model settings and regulatory requirements all influence the results.

For projects involving complex terrain, sparse weather observations or potentially significant air quality impacts, the difference between a routine model setup and a defensible assessment can be substantial.

An AERMOD model may be appropriate for many regulatory applications, while CALMET and CALPUFF offer capabilities suited to certain complex meteorological conditions and longer-range transport problems. Choosing the right approach requires expert judgement, an understanding of the site, the modelling objective and the applicable regulatory guidance.

What do CALMET and CALPUFF do?

CALMET is a three-dimensional meteorological model. It processes meteorological observations and other information to develop wind and temperature fields across a modelling domain. CALPUFF uses those fields to simulate the transport and dispersion of pollutants released by industrial sources.

We're taking care of it for you

CALMET combines an initial estimate of the meteorological field with available observations and adjustments related to terrain and other physical influences. Its diagnostic wind-field calculations involve expert judgement and decisions and about how the available information should be represented.

CALPUFF then uses the resulting meteorological fields alongside emissions, source characteristics and other inputs to predict pollutant concentrations.

Both models have numerous settings. Many can be retained at their recommended defaults when those choices suit the application. Others deserve closer examination when site conditions, data limitations or the modelling objective make the default assumptions questionable.

The key question is whether the selected configuration adequately represents the conditions relevant to the assessment.

When should a modeller use expert judgement to question the defaults?

Expert judgement is especially important when the available data or the physical environment challenges the assumptions behind a standard model setup.

Examples include:

• Complex terrain: Valleys, ridges and steep slopes can influence local airflow in ways that regional weather observations may not capture.
• Sparse meteorological observations: A distant weather station may not represent wind conditions at the facility or nearby receptors.
• Stable atmospheric conditions: Temperature inversions and weak winds can affect plume behaviour and ground-level concentrations.
• Coastal or lake environments: Local circulations and shoreline effects may require specialized consideration.
• Potentially consequential results: A small change in predicted concentration may affect the interpretation of compliance with an applicable air quality objective.

A modeller should first identify the uncertainty or physical process that needs attention. The next step is to determine whether the model represents that process adequately, what evidence is available and whether alternative assumptions warrant testing.

Changing a setting simply because it is available does not reflect expert judgement and is not a sound modelling strategy. Every material change should have a technical rationale.

CALMET: Building a defensible wind field

CALMET offers options that influence how meteorological information is processed. Their relevance depends on the site, the input data and the intended application.

CALMET parameters such as BIAS, IEXTRP, RMAX1–RMAX3, RMIN, R1 and R2 can affect aspects of meteorological processing. Their meanings and appropriate use must be checked against the documentation for the specific software version.

For example, a modeller assessing a facility in a narrow valley should use expert judgement to consider whether the available observations represent the local wind field. A distant upper-air station and a surface station outside the valley may provide useful information, but neither necessarily captures all the local effects.

The appropriate response might involve reviewing the observations, examining terrain data, investigating alternative configurations or obtaining additional meteorological measurements. The decision should follow from the evidence rather than a blanket rule about which observations deserve more weight.

Meanwhile, an AERMOD assessment has its own meteorological processing requirements and assumptions. The same general principle applies across modelling systems: data suitability and model configuration must be evaluated with expert judgement in the context of the actual site.

A clearer view of potential impacts and more accurate predictions

CALPUFF: Selecting dispersion options

After the meteorological fields have been developed, CALPUFF settings determine how the model represents pollutant transport and dispersion.

Options such as MSLUG, MSPLIT and MDISP relate to different aspects of puff representation and dispersion calculations. Other options address specialized physical effects, including certain subgrid-scale terrain or shoreline processes.

The appropriate choice depends on the modelling objective, the physical conditions and the software version. A long-range transport assessment, for example, may require expert judgement in choosing from different considerations from an assessment focused on nearby receptors.

Three questions should guide the decision:

1. Does the selected option represent a process that matters for this application?
2. Are the required input data and assumptions defensible?
3. Is there evidence that an alternative configuration would materially affect the results?

The answer may be to retain the standard setting. In other cases, the modeller may need to investigate alternatives and document why one configuration is more appropriate.

Specialized chemistry options require additional care. Chemical transformation and secondary pollutant formation may depend on background concentrations and other assumptions. With the correct type of expert judgement, the appropriate treatment could be established from the assessment requirements and applicable guidance.

Example: CALMET in Alberta foothills terrain

Illustrative scenario — not a Calvin Consulting project. The site details, model configurations and numerical outcomes below require confirmation before being presented as actual project evidence.

The question - Consider a proposed industrial facility in a river valley near the Alberta foothills. An air quality assessment is required to evaluate predicted ground-level concentrations of sulphur dioxide (SO₂) and other relevant pollutants.

The results will help inform the project's environmental assessment and regulatory application.

The challenge - Valley walls and surrounding foothills may influence local wind direction, drainage flows and atmospheric stability. Regional meteorological stations may be several tens of kilometres away and may not fully represent the site's local conditions.

The modelling question is whether the meteorological data and CALMET configuration adequately represent the airflow that governs pollutant transport toward nearby receptors.

The baseline and alternative - A baseline configuration could use suitable regional meteorological observations, available terrain and land-use data and an initially selected CALMET grid.

The modeller would then assess whether the baseline adequately represents the relevant site conditions. Depending on the evidence, alternatives might include a different grid resolution, a revised treatment of terrain effects or the use of suitable additional meteorological observations.

The exact options must be selected from the capabilities of the applicable model version. Specialized settings should only be used where their assumptions and data requirements are satisfied.

How should the alternatives be evaluated? A defensible comparison would examine:

• The wind fields produced by each configuration under representative conditions.
• Predicted concentrations at relevant receptors and across the modelling domain.
• Changes in the location and magnitude of predicted maximum concentrations.
• The implications for applicable air quality objectives.
• The sensitivity of the results to the assumptions being investigated.

What did the comparison establish? A material change in predicted concentrations could indicate that the original configuration did not adequately represent an influential process. However, higher concentrations alone do not prove that an alternative is more accurate. The modeller must deploy his expert judgement to evaluate whether the alternative is better supported by the available evidence.

The final configuration should be selected on technical grounds and the rationale documented in the assessment. Remaining uncertainty should also be acknowledged.

This is where professional judgement provides practical value: it helps establish which assumptions matter, which alternatives deserve investigation and what the results can reasonably support.

Accuracy, avoiding errors, and minimizing risks

Sensitivity testing, uncertainty and regulatory defensibility

A technically correct model run can still produce an unreliable assessment if expert judgement is overlooked and important inputs or assumptions are poorly chosen.

Sensitivity testing helps determine whether plausible changes to influential inputs or settings affect predicted concentrations, spatial patterns or the interpretation of potential impacts. It also helps distinguish decisions that materially affect the assessment from those that make little practical difference.

A defensible modelling report should explain:

• Why the selected modelling system and configuration are appropriate.
• Which meteorological data and terrain information were used.
• Why important assumptions and non-default settings were selected.
• What sensitivity tests were conducted and what they showed.
• Which limitations remain and how they affect interpretation.

For Alberta regulatory applications, consult the current Air Quality Model Guideline and related modelling guidance. Alberta's published guidance includes the 2021 Air Quality Model Guideline and supplemental guidance for interpreting regulatory air quality model assessments effective July 14, 2025. Confirm the current applicable documents and requirements when preparing a new assessment.

Predicted concentrations must be compared against the appropriate Alberta Ambient Air Quality Objectives and Guidelines, including the applicable averaging periods and statistical forms. The correct comparison depends on the pollutant, regulatory context and current requirements.

An assessment using AERMOD also requires appropriate inputs, processing choices and regulatory interpretation. The choice between AERMOD and CALPUFF should follow the applicable guidance and the technical needs of the project, rather than personal preference.

Get the modelling right before it becomes an operating problem

Air dispersion modelling assessment carried out with appropriate expert judgement can influence facility design, emissions controls, operating decisions and regulatory approvals. An unsuitable assumption can lead to misleading conclusions, unnecessary mitigation or additional work when a regulator questions the assessment.

Specialist support helps identify the important decisions early, select appropriate methods and focus analysis where it can change the outcome. It can also avoid spending time and money on unnecessary model complexity.

Calvin Consulting Group Ltd. provides air quality consulting and dispersion modelling support for industrial projects. We help clients evaluate emissions, meteorological data, modelling assumptions and predicted impacts with a focus on technically defensible results.

If your project requires an AERMOD assessment or a CALMET/CALPUFF study, contact Calvin Consulting to discuss the site, the regulatory requirements and the modelling approach appropriate to your needs.

The value of expert modelling is a clear, defensible answer to the question your project needs to resolve.

If you are looking for a good group of modelers for your project, please contact Calvin Consulting Group Ltd.  Calvin Consulting Group Ltd. has been around for more than 20 years, and its team is made up of experienced professionals.  Besides that, they've done a lot of successful projects.

Go back from Expert Judgement to the Air Quality Testers web page, or visit the Stuff in the Air homepage.

A practical checklist for site operators

Before accepting an air dispersion modelling assessment, ask:

• Is the selected model appropriate for the site and regulatory application?
• Are the meteorological data representative of the facility and surrounding area?
• Have terrain and other relevant physical influences been considered?
• Are important model settings supported by technical reasoning?
• Were material uncertainties investigated through sensitivity testing where appropriate?
• Are the predicted concentrations compared against the correct regulatory criteria?
• Can the consultant explain the limitations of the results and defend the final configuration?

These questions help distinguish a model run from a defensible air quality assessment.

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.