View all publications

1 Introduction

1.1 Aims and objectives

The purpose of this Guide is to provide good practice protocols for carrying out atmospheric dispersion modelling in New Zealand. Where the recommended protocols are not suitable for the particular modelling exercise, the reasons for deviating from them should be clearly explained. In establishing these good practice protocols, the guide aims to improve the use of models in New Zealand and consequently the accuracy of modelling results so they can be relied upon when considering the potential adverse effects of a discharge to air.

The Guide contains information and recommended protocols on many aspects of modelling including: the main types of model available and when to use them, the nature of input data required, and how to get the most accurate results for the level of assessment required. It is designed to assist those relatively new to modelling who may have taken a course or two, and those involved in reviewing modelling outputs for auditing resource consent applications. The 'recommended protocol' shaded boxes should also be useful for expert modellers who are seeking better consistency in how models are used in New Zealand.

It should be recognised that modelling is a complex process and that some training in the form of workshops or courses is advisable before commencing modelling. This Guide will assist in recalling the training you have received and it sets specific protocols to follow where alternative options are available.

The Guide focuses on how to get accurate data once the decision to model has been made. Guidance on when to model and how to interpret modelling results, in terms of evaluating the potential effects of the discharge on the environment, will be contained in a separate document currently being prepared by the Ministry entitled the Good Practice Guide for Assessing Discharges to Air. Although these two areas are integrally linked they have been separated to avoid excessive complexity in one document. However, both guidance documents should be reviewed when assessing a discharge to air using dispersion modelling.

Once the decision to model has been made, the Guide can help practitioners to determine:

  • which model is most appropriate for the particular circumstances
  • what data to put into the model (including emissions data and meteorological data)
  • how to run a model effectively
  • pitfalls to watch out for
  • how to understand the accuracy of modelling results.

The Guide also discusses the advantages and limitations of:

  • current practice associated with using steady-state dispersion models as an assessment tool
  • new generation models.

The Guide mainly covers the use of dispersion models to assess the effects of pollutants discharged from point (and multiple point) sources. However, modelling of area and line sources is also briefly considered.

Throughout the Guide modellers are encouraged to:

  • use the best available information
  • comply with the recommendations made in this document and consider applying the guidance
  • create an auditable trail of the work undertaken.

The guidance is not intended to replace the detailed user manuals that accompany each dispersion model and these should still be consulted. In addition it should be recognised that the recommendations do not have any regulatory status and they can be deviated from as required and when justified. Neither is the advice in any way government policy.

1.2 Overview

Here is an outline of the information contained in the Guide.

Section 1 (this section) presents the background information that puts atmospheric dispersion models into a wider context and highlights the issues that should be considered before using them.

Section 2 contains a brief review of the Gaussian-plume and advanced models that are commonly used for regulatory applications.

Section 3 contains a brief review of the more specialised applications of dispersion modelling.

Section 4 details processes for determining information that should go into an atmospheric dispersion model to ensure good quality information is obtained.

Section 5 details the importance of, and methods for, acquiring reliable and representative meteorological input for air quality modelling purposes. The meteorological requirements of advanced dispersion models are described and methods by which these requirements can be met are discussed.

Section 6 describes how to present and explain modelling results clearly and simply, including the interpretation of modelling results and addressing the uncertainty in model predictions.

1.3 What is an atmospheric dispersion model?

A model is a simplified picture of reality. It doesn't contain all the features of the real system but contains the features of interest for the management issue or scientific problem we wish to solve by its use. Models are widely used in science to make predictions and/or to solve problems, and are often used to identify the best solutions for the management of specific environmental problems.

Models may be:

  • physical - a scaled-down representation of reality
  • mathematical - a description of the system using mathematical relationships and equations.

Contaminants discharged into the air are transported over long distances by large-scale air-flows and dispersed by small-scale air-flows or turbulence, which mix contaminants with clean air. This dispersion by the wind is a very complex process due to the presence of different-sized eddies in atmospheric flow. Even under ideal conditions in a laboratory the dynamics of turbulence and turbulent diffusion are some of the most difficult in fluid mechanics to model. There is no complete theory that describes the relationship between ambient concentrations of air pollutants and the causative meteorological factors and processes.

An atmospheric dispersion model is a:

  • mathematical simulation of the physics and chemistry governing the transport, dispersion and transformation of pollutants in the atmosphere
  • means of estimating downwind air pollution concentrations given information about the pollutant emissions and nature of the atmosphere.

Dispersion models can take many forms. The simplest are provided in the form of graphs, tables or formulae on paper. Today dispersion models more commonly take the form of computer programs, with user-friendly interfaces and online help facilities.

Most modern air pollution models are computer programs that calculate the pollutant concentration downwind of a source using information on the:

  • contaminant emission rate
  • characteristics of the emission source
  • local topography
  • meteorology of the area
  • ambient or background concentrations of pollutant.

A generic overview of how this information is used in a computer-based air pollution model is shown in Figure 1.1.

Figure 1.1: Overview of the air pollution modelling procedure

The process of air pollution modelling contains four stages (data input, dispersion calculations, deriving concentrations, and analysis). The accuracy and uncertainty of each stage must be known and evaluated to ensure a reliable assessment of the significance of any potential adverse effects.

Currently, the most commonly used dispersion models are steady-state Gaussian-plume models. These are based on mathematical approximation of plume behaviour and are the easiest models to use. They incorporate a simplistic description of the dispersion process, and some fundamental assumptions are made that may not accurately reflect reality. However, even with these limitations, this type of model can provide reasonable results when used appropriately.

More recently, better ways of describing the spatially varying turbulence and diffusion characteristics within the atmosphere have been developed. The new generation dispersion models adopt a more sophisticated approach to describing diffusion and dispersion using the fundamental properties of the atmosphere rather than relying on general mathematical approximation. This enables better treatment of difficult situations such as complex terrain and long-distance transport.

Sections 2 and 3 provide detailed descriptions of the different dispersion models available, what each model can potentially be used for, and their benefits and problems.

1.4 The importance of meteorology

The ground-level concentrations resulting from a constant discharge of contaminants change according to the weather (particularly the wind) conditions at the time. Meteorology is fundamental for the dispersion of pollutants because it is the primary factor determining the diluting effect of the atmosphere. Therefore, it is important that meteorology is carefully considered when modelling.

The importance of, and methods for, acquiring reliable and representative meteorological input for air quality modelling purposes are detailed in section 5.

1.5 What can dispersion modelling be used for?

Models can be set up to estimate downwind concentrations of contaminants over varying averaging periods - either short term (three minutes) or long term (annual). In New Zealand, the most common use of dispersion modelling is to assess the potential environmental and health effects of discharges to air from industrial or trade premises. Such assessments are required to be undertaken in accordance with the Resource Management Act 1991 (RMA) for applications for discharge permits. Models are particularly valuable for assessing the impacts of discharges from new activities and to estimate likely changes as a result of process modifications.

Modelling results can also be used for:

  • assessing compliance of emissions with air quality guidelines, criteria and standards
  • planning new facilities
  • determining appropriate stack heights
  • managing existing emissions
  • designing ambient air monitoring networks
  • identifying the main contributors to existing air pollution problems
  • evaluating policy and mitigation strategies (e.g. the effect of emission standards)
  • forecasting pollution episodes
  • assessing the risks of and planning for the management of rare events such as accidental hazardous substance releases
  • estimating the influence of geophysical factors on dispersion (e.g. terrain elevation, presence of water bodies and land use)
  • running 'numerical laboratories' for scientific research involving experiments that would otherwise be too costly in the real world (e.g. tracking accidental hazardous substance releases, including foot-and-mouth disease)
  • saving cost and time over monitoring - modelling costs are a fraction of monitoring costs and a simulation of annual or multi-year periods may only take a few weeks to assess.

1.6 What can't dispersion models do?

Even the most sophisticated atmospheric dispersion model cannot predict the precise location, magnitude and timing of ground-level concentrations with 100% accuracy. However, most models used today (especially the US EPA approved models) have been through a thorough model evaluation process and the modelling results are reasonably accurate, provided an appropriate model and input data are used.

Errors are introduced into results by the inherent uncertainty associated with the physics and formulation used to model dispersion, and by imprecise input parameters, such as emission and meteorological data. The most significant factors that determine the quality and accuracy of the results are:

  • the suitability of the model for the task
  • the availability of accurate source information
  • the availability of accurate meteorological data.

The causes of model uncertainty and the methods by which they should be addressed when using dispersion models are discussed in more detail in section 6.2.

1.7 When is it appropriate to use dispersion modelling as an assessment tool?

Atmospheric dispersion models may not always be the most appropriate method for assessing the potential environmental impacts of a discharge to air. Guidance on when modelling is required as part of an assessment of environmental effects will be covered in more detail in the Good Practice Guide for Assessing Discharges to Air currently under development by the Ministry.

Modelling is unlikely to be needed when a discharge is already permitted by a regional plan. However, councils may specify when modelling is required for particular activities. Assessors should consult with relevant councils to determine whether modelling is required before commencing assessments and submitting applications.

Recommendation 1

Before undertaking an assessment of effects using atmospheric dispersion modelling, the proposed approach for assessing adverse effects should be discussed with the relevant council (national guidance will be covered in the Good Practice Guide for Assessing Discharges to Air currently under development by the Ministry).

Alternative and perhaps more pragmatic methods of providing information to support assessments should be employed when the scale of the activity is small and its potential environmental effects are likely to be minor, or when modelling is unlikely to provide good-quality information.

Atmospheric dispersion models should only be used when they are appropriate for investigating the scale and significance of the effects of a discharge on the environment, and their use should be justified.

Users must recognise that there are limitations to the scope of a model's application and to the accuracy of model predictions. These should be identified and discussed in conjunction with the modelling results.

Modelling results provide reasonably accurate predictions of ground-level concentrations of contaminants from a discharge, provided input parameters are appropriate. Factors influencing their accuracy should be estimated, reported and acknowledged.