Air Quality Modelling

Modelling refers to a set of scientific methods that are used to represent the complicated processes that govern meteorology and the behaviour of pollutants in the atmosphere. Modelling is used to identify the sources that contribute to poor air quality and to project air quality changes for different "what if" scenarios: e.g., what would happen to the air quality if a new highway, industrial or housing development were built? In this way, modelling helps to inform decisions on how to maintain and improve air quality.

There are hundreds of different kinds of models used for air quality management. Some are very simple, others are very complex. Some common types of models include:

  • Dispersion models — which use equations to represent the way that pollutants travel in the air in order to calculate the downwind air concentrations that result from emitting something into the air;
  • Receptor models — where the properties of the air pollutants are able to identify and quantify the sources of air pollutants;
  • Meteorological models — which use equations that represent the behaviour of the atmosphere, in order to predict the meteorological conditions at specific locations and forecast what these conditions will be in the future;
  • Physical models — small scale reproductions used in wind tunnels to simulate actual conditions; and
  • Statistical models — where statistics are used to provide a link between emissions and the resulting concentrations.

The provincial government's involvement in air quality modelling includes:

  • conducting modelling to support air quality management decisions;
  • providing modelling guidance;
  • conducting reviews and making decisions on the acceptability of dispersion modelling submitted by consultants to the Ministry of Environment in support of permit applications and environmental assessment; and
  • collecting meteorological and air quality data in support of air quality modelling.

Dispersion Modelling

What is an Air Quality Dispersion Model?

An air quality dispersion model is a series of equations that mathematically describe the behaviour of pollutants in the air. It provides a cause-effect link between the emissions into the air and the resulting air pollution concentrations. Dispersion models have been used in many different applications, but have traditionally been used for air quality assessments in support of decisions regarding approvals and permits for regulated sources.

This simple but powerful ability means that dispersion models can be used to examine scenarios that would otherwise be too expensive, difficult or destructive to do in the real world. For example, models can be used to answer difficult questions such as:

Which emission sources are causing the air quality to be poor?
A model can identify sources that are the greatest contributors to poor air quality, and thus provide air quality managers some idea of which sources should receive the highest priority to control.

Should a proposed factory be allowed to be built?
A model can determine whether a proposed new source will cause an air quality problem, and thus help in decisions on project approval.

How close should a residential area be allowed to encroach on a chemical refinery?
A model can delineate risk zones from an uncontrolled release of a poisonous gas and thus help in land use and planning decisions.

How Dispersion Models Work

Dispersion models are based on our understanding of how pollutants travel and disperse in the air. Equations, which represent the science of this behaviour, are used to calculate the downwind air concentrations that result from a source that emits something into the air.

Some models have basic equations, and other more complex models include more detailed equations. However, all models are a representation of reality (not reality itself); we cannot expect a model to provide perfect results. In some situations the model may predict the actual concentrations very well, and in others it may not.

The reasons for this are many. Since our understanding of the details of the behaviour of air pollution is not complete, it follows that the equations which describe this process are not complete. In addition, models require inputs (emissions, weather and geography). If these inputs are of poor quality, then the model output will also be of poor quality (i.e. the model will not magically correct poor input).

Weather information is a critical model input. This is available from the many meteorological stations operated throughout the province by the Ministry of Environment, as well from other sources such as Environment Canada airport sites.

Interpreting the output and assessing how much confidence decision-makers should place in the results involves considerable expertise and experience. Those involved in dispersion modelling need to make sure that the correct model is selected for a given situation, that all the inputs have been checked for completeness and accuracy, and that testing has been done to ensure that the model has be used correctly. Furthermore, modellers need to be skilled in providing guidance to decision-makers in how the output should be used.

Air quality data collected by the Ministry of Environment is used to determine how well a model performs by comparing model predictions to actual observations. This provides some level of confidence in the model, while helping to identify any parts of the model that need to be improved.

Guidelines for Dispersion Modelling

Given the importance of dispersion models and their growing use, the Government of British Columbia has developed the British Columbia Air Quality Dispersion Modelling Guideline (PDF) in consultation with dispersion modelling experts. The Guideline provides recommendations on models accepted for use in B.C. and how they should be applied.

The Dispersion Modelling Plan (DOC) can be filled in electronically and e-mailed. Alternatively, it can be printed out and filled in manually, then faxed or mailed to the ministry.

This will ultimately help in promoting consistency and establishing good practices so that models will be applied in a way that provides the information needed to inform decision-making. A primer is also available which is intended to help non-experts understand the key messages of the Guideline: A Primer on the British Columbia Air Quality Dispersion Modelling Guideline (PDF).


Receptor Modelling

Monitoring air quality can tell us a lot about the risk to local human health and it can tell us about seasonal, daily or even hourly conditions. However, by itself, monitoring can not tell us where the pollution is coming from or what the relative contribution from different emission sources is. Common emissions sources include wood stoves, forest fires, vehicles, industrial stacks and road dust.

To help air quality planners and decision-makers determine the sources behind bad air quality, and develop plans to reduce air pollution, scientists use a variety of source apportionment tools. In air pollution research, source apportionment is the task of determining sources of air pollutants and assigning or "apportioning" the relative contribution of each unique source to overall air quality.

A traditional approach is to use dispersion modelling where a pollutant emission rate and meteorological information are input to a mathematical model that "disperses" (and sometimes, chemically transforms) the pollutant, and predicts into the future what the air quality will be like at a certain place and a certain time.

A speciation monitor collecting air quality samples

A traditional approach is to use dispersion modelling where a pollutant emission rate and meteorological information are input to a mathematical model that "disperses" (and sometimes, chemically transforms) the pollutant, and predicts into the future what the air quality will be like at a certain place and a certain time.

Receptor modelling is a newer approach to source apportionment and has been developed using more and more complex models over the last 30 years. Unlike dispersion modelling, receptor modelling cannot predict future air quality but, instead, looks at past data collected at one site over a specific time period to determine the sources to that site.

Data for the model is collected using a speciation sampler (pictured above), which takes integrated samples over a 24-hour period from ambient air. These samples are analysed by a lab for a suite of chemical components.

The speciated data is then plugged into a computer model. The output of the model provides groupings of chemical species which are then interpolated as emission sources. The end result of the models resembles a "recipe" of air pollutant sources, and their respective contribution to the overall air quality at the site.

Current Receptor Modelling in British Columbia

There are three active speciation monitors running in British Columbia: Quesnel, Abbotsford and Burnaby. Receptor modelling has been performed in several B.C. communities including: Golden, Prince George, Kelowna and Burnaby. Reports on the Golden and Prince George studies are available in the Reports and Publications section.


Meteorological Modelling

Computers can be used to solve equations that describe the current and expected behaviour of the atmosphere, and to produce current and forecast meteorological information over large geographic areas.

These sets of equations are called "meteorological models" and typically require powerful computers to produce output on winds, temperature, and precipitation at different locations and vertical levels in the atmosphere. Such models account for the effects of topography, oceans and global circulation patterns.

Observations of meteorological conditions are taken at specific locations across the province, and around the world. Due to the complex topography in British Columbia and the large distances between observation sites, determining the precise meteorological conditions for areas where there are no observations is nearly impossible.

However, by applying a meteorological model, the conditions at any location can be estimated. In addition, more sophisticated versions of these models can forecast the weather for any location and such tools are used to produce weather forecasts that are relied upon by the public and private sector.

From an air quality perspective, the meteorology determines the transport and dispersion of pollutants. Thus these meteorological models can be used to help estimate the current and forecast location and concentrations of pollutants that result from emission sources. In other words, the output from these models can be a critical input into dispersion models.