Air Pollution Climatology

Air pollution is defined as an atmospheric condition in which substances (air pollutants) are present at concentrations higher than their normal ambient (clean atmosphere) levels to produce measurable adverse effects on humans, animals, vegetation, or materials (Seinfeld, 1986). Polluting substances can be noxious or benign, and can be released by natural and anthropogenic (human-made) sources. According to the World Health Organization an estimated 3 million people die each year because of exposure to air pollution (WHO, 2000). Air pollution climatology is concerned with the study of atmospheric phenomena and conditions that lead to occurrence of large concentrations of air pollutants and with their effects on the environment.

Air pollutants are typically classified into three categories: suspended particulate matter (SPM), gaseous pollutants (gases and vapors) and odors. SPM in the air includes PM₁₀ (particulate matter with median diameter less than 10 µm), PM_2.5 (particulate matter with median diameter less than 2.5µm), diesel exhaust, coal fly-ash, mineral dusts (e.g. asbestos, limestone, cement), paint pigments, carbon black and many others. Gaseous pollutants include sulfur compounds (e.g. sulfur dioxide (SO₂)), nitrogen compounds (e.g. nitric oxide (NO), ammonia (NH₃)), organic compounds (e.g. hydrocarbons (HC), volatile organic compounds (VOC) and polycyclic aromatic compounds (PAH), etc.). Known odorous agents are generally sulfur compounds such as hydrogen sulfide (H₂S), carbon disulfide (CS₂) and mercaptans.

The past few decades have seen much progress in the understanding of air pollution climatology and meteorology, due to rapid progress in pollutant measurement techniques, increased understanding of atmospheric dynamics, and computational capabilities, particularly for areas with complex terrain. At the time of the last edition of this Encyclopaedia (1986), the processes of air pollution dispersion were well understood for

Table A11 Typical spatial and temporal scales of air pollution problems

urban areas over relatively simple terrain. However, knowledge gaps existed regarding the dispersion of pollutants in cities near coastlines or lake shores, or in areas of complex topography such as mountainous areas — all of which may be considered as “complex terrain”. In addition, air pollution climatology now considers not only localized issues, but also global problems such as stratospheric ozone reduction. The role of long-range transport of pollutants in degradation of air quality in remote/distant regions has also been recognized. Over the past few decades a combination of field observational studies — such as the European Tracer Experiment (ETEX; Girardi et al., 1998) and the Across North America Tracer Experiment (ANATEX; Draxler et al., 1991) — in conjunction with different modeling approaches — has added to a growing knowledge base concerning the complexities of air pollution dispersion.

Natural air pollution has always occurred throughout Earth’s history. Volcanic eruptions, natural fires and wind-blown dust are examples of natural phenomena that introduce pollutants into the atmosphere. Anthropogenic sources have only become a serious problem during the past 200 years with the growing population, increased urbanization and accelerating industrialization. It is estimated that 2 billion metric tons of air pollutants are released into the atmosphere world-wide (Arya, 1999). Air quality is affected not only by emissions into the atmosphere, but also by the pollution potential of air. Stern et al. (1984) provide a useful classification for the types and scales of air pollution problems (Table A11).

Earlier examples of important literature on air pollution climatology include Slade (1968), Williamson (1972), Stern (1976), Holzworth (1974) and Pasquill (1974). However, several introductory books that contain in-depth information on recent developments in this discipline are now available. Arya (1999) is an invaluable book that covers the physics of meteorology and dispersion, while Jacobson (2002) includes perspectives on history and science. The book by Seinfeld and Pandis (1998) also contains a wide range of topics on the physics and chemistry of air pollution. Those who require an extensive treatment of atmospheric chemistry (ambient and polluted) are encouraged to consult Finlayson-Pitts and Pitts (1986). A concise literature review is provided by Sturman (2000).

This account will focus on air pollution climatology by first introducing contemporary air pollution problems and issues; then by considering the components of air pollution, including emissions, atmospheric conditions and receptor response. The item continues by discussing air pollution dispersion models and air quality guidelines and standards.

Contemporary air pollution problems and issues Air quality issues came to the forefront of scientific and societal attention during the 1950s. In 1952 London’s “killer fog” caused a great increase in human fatality (4000 excess deaths), especially in people with a history of cardiopulmonary problems. The episode resulted in the implementation of air pollution mitigation measures by the authorities (Brimblecombe, 1987). The air pollution in London was due to the emission of smoke from the burning of coal and other raw materials into the foggy atmosphere; this type of pollution is known as the London-type smog (smoke plus fog). In the same decade another type of smog, which is formed by chemical reactions in the atmosphere — the Los Angeles photochemical smog — also gained notoriety. During the 1970s acid rain emerged as the top environmental concern. A decade later the reduction of stratospheric ozone due to emission of anthropogenic compounds became the top international concern.

Major air pollution problems facing humanity in the 21st century include urban smog, acid deposition, indoor air pollution, Antarctic stratospheric ozone depletion (global stratospheric ozone reduction), and the highly contentious issue of global warming due to alteration of the global longwave (infrared) radiation balance caused by emission of greenhouse gases (GHG) such as carbon dioxide, methane, and nitrous oxide. Table A12 lists some of the polluting substances present in the atmosphere.

Urban smog

This is characterized by build-up of harmful gases and particulates emitted from vehicles, industry, or other sources (primary pollutants) — that cause the London-type smog, or is formed chemically in the air from emitted precursors (producing secondary pollutants) — that cause the Los Angeles-type smog. London-type smog events have been observed in various places around the world, such as the Meuse Valley in Belgium, Pittsburgh in the United States, and Christchurch in New Zealand, to name just a few. Photochemical smog has been observed in most cities around the world; the problem is particularly severe in mega-cities such as Mexico City, Tokyo, Beijing and Tehran.

Transboundary pollution

This is characterized by transport of pollutants across political boundaries and/or vast distances. Examples of such pollution include acid deposition (also known as acid rain — a term introduced by Robert Angus Smith in the nineteenth century) and regional haze caused by forest fires. Acid deposition happens when sulfuric acid (H₂SO₄), hydrochloric acid (HCl), or nitric acid (HNO₃) is deposited on the ground in vapor form or dissolved in rainwater, fogwater, or particulates. This increased acidity in turn harms soils, lakes, vegetation and materials. Cowling (1982) provides an historical perspective on the discovery of and mitigation measures used to deal with acid rain. Natural examples of transboundary pollution include the 1998 dust clouds caused by intense storm events in western China, which carried vast quantities of particulate matter across the Pacific Ocean, contributing to increased concentrations in Vancouver, Canada (McKendry et al., 2001).

Indoor air pollution

Most people spend a large amount of time indoors where they are constantly exposed to indoor air pollution. This is caused from either the emission of gases and particulates in enclosed

Table A12 Gases and particulate matter components in air pollution

buildings or transport of pollutants from outdoors; but it is not as extensively researched as outdoor air pollution (Wanner, 1993). Health effects of indoor air pollutants can be severe and directly affect the respiratory and cardiovascular systems. Indoor air pollutants include radon gas, ammonia, volatile organic compounds (VOC), polycyclic aromatic hydrocarbons (PAH) and second-hand cigarette smoke. Radon gas has a long-term health effect and is naturally released by soil due to radioactive decay of radium (Bridgman et al., 2000). Indoor exposure to particulate matter increases the risk of acute respiratory infections, leading to an increase in infant and child mortality rates in developing countries. WHO reports that such exposure is responsible for between half and one million excess deaths in Asia, and 300 000–500 000 excess deaths in sub-Saharan Africa (WHO, 2000).

Antarctic ozone depletion

Antarctic ozone depletion and the reduction of global stratospheric ozone are due to emission of chlorine and bromine compounds such as chlorofluorocarbons (CFC). CFC break down by photolysis reaction (breakdown of molecules by solar radiation) after they have traveled to the upper atmosphere. The reduction of ozone increases the amount of ultraviolet (UV) radiation that reaches Earth’s surface. UV radiation can damage genetic material of organisms and in humans it is known to cause skin cancer.

Scientific consensus supports the theory that human emission of carbon dioxide, methane, nitrous oxide and other gases may play a significant role in the cause of global warming (IPCC, 2001). Therefore global warming is also an air pollution issue.

Components of an air pollution problem

Air pollution problems usually have three components: (1) emission of polluting substances into the air; (2) the pollution potential of the atmosphere characterized by its ability to transport, diffuse, chemically transform and remove the pollutants; and (3) response of the receptors (e.g. people, animals, vegetation) to the exposed concentration. Each of these components is considered in this section.

Pollution emissions

Emission sources can be categorized as follows:

• 1. Urban and industrial. Industrial sources include power generation plants that use fossil fuel, mining, manufacturing, smelting, pulp and paper plants, and chemical industries. Major gaseous pollutants emitted are carbon monoxide (CO), carbon dioxide (CO₂), sulfur dioxide (SO₂) and volatile organic compounds.

• 2. Agricultural and rural. Agricultural areas can be sources of air contaminants due to decaying waste from animals and plants which can release ammonia, methane (CH₄, a powerful greenhouse gas) and other noxious gases. In addition, windblown dust due to plowing, tilting, and harvesting, and smoke and haze due to slash burning can lead to severe degradation in air quality. An interesting example of air pollution caused by agricultural practices is the regional haze and smog events that plagued the Southeast Asian countries in 1997 due to forest fires in Indonesia (Davies and Unam, 1999; Khandekar et al., 2000). Burning forests is a traditional way of clearing land for agricultural purposes in Indonesia (and indeed in many other parts of the world), but the fires became unmanageable due to prolonged drought and the delayed onset of the monsoon rains caused by the 1997 El Niño episode. A strong increase in tropospheric ozone concentrations occurred over tropical Southeast Asia, reaching as far as Hong Kong (Chan et al., 2001).

• 3. Natural. In addition to anthropogenic sources, many natural sources can contribute to air pollution. These include: volcanic eruptions that release vast quantities of particulate matter, CO₂, SO₂, and other gases, into the atmosphere; biogenic (from biological sources) emissions from forests and marshlands of compounds such as hydrocarbons (terpenes and isoprenes; Guenther et al., 1995), methane and ammonia; and soil microbial processes which contribute NO, CH₄ and H₂S.

Atmospheric conditions

The state of the atmosphere determines its pollution potential, which is affected by three important processes: dispersion which is the horizontal and vertical spread and movement of pollutants; transformation which involves chemical reactions that occur between pollutants or to pollutants under certain temperature and sunlight conditions; and removal of pollutants through such mechanisms as dry and wet deposition (Figure A15). Each of these processes will be considered in turn, but first the overriding control of synoptic conditions will be considered.

Figure A15

Schematic illustration of emission, transmission and deposition of air pollution.

Synoptic controls

“Synoptic conditions” refers to an intermediate scale of atmospheric activity between global and regional scales and includes circulation systems such as anticyclones and cyclones. Synoptic circulation features have a spatial dimension of the order of thousands of kilometers and may last about a week on average. Atmospheric circulation at this scale has an important control on air pollution since it influences cloud (amount, thickness, height and type), temperature and relative humidity of the airmass, the type and amount of precipitation, and the wind speed and direction. These parameters in turn influence the vertical temperature structure of the atmosphere which determines the atmospheric stability. If the air temperature decreases rapidly with height (i.e. at a rate greater than the dry adiabatic lapse rate of 9.8°C km⁻¹) then the atmosphere is unstable. This means that vertical motion is promoted and any emissions will mix well into the boundary layer (the lowest layer of the atmosphere that is affected by the diurnal cycle of temperature, humidity, wind, and pollution). However, in stable conditions, which occur when temperature either decreases slowly with height (at a rate less than the dry adiabatic lapse rate) or increases with height (which is known as a temperature inversion), pollution may become trapped close to the emission source, resulting in increased concentrations. The layer through which the mixing of pollutants occurs is known as the mixed layer, and the height to which pollutants mix is called the mixing depth or mixing height.

Anticyclonic circulation is more conducive to pollution episodes since it usually involves weaker synoptic winds and clear skies. In summer, stronger surface heating under anticyclonic flow results in warmer temperatures in the boundary layer, and this warmth, together with high solar radiation levels and generally calm or light airflow, can lead to the formation of photochemical smog (discussed in more detail later). Since anticyclonic circulation involves subsidence of air towards the surface (and hence warming), this results in elevated inversions which can trap pollutants in a shallow mixed layer (Figure A16a).

Figure A16

Air pollutants emitted at the surface can become trapped by (a) an elevated inversion layer and (b) a surface inversion layer (modified after Whiteman, 2000).

In winter, clear skies and light winds during anticyclonic conditions lead to strong nocturnal radiative cooling of the surface which results in surface or radiative inversions (Figure A16b). During such conditions any pollutants emitted at the surface (e.g. through industrial emissions, domestic heating or traffic) become trapped very close to the surface and may result in high pollutant concentrations. This is the typical scenario that produces classic smog as a combination of smoke and fog. Figure A17 shows the severe degradation in air quality in Christchurch, New Zealand, as a result of emission of smoke from solid-fuel burning (wood) into a shallow surface inversion layer.

Figure A17

Christchurch, New Zealand, on a winter morning. The thick smoke at the surface is the result of poor ventilation due to a surface inversion layer that formed overnight. Smoke and particulate matter is emitted from home heating with solid-fuel burners (courtesy of The Christchurch Press).

Dispersion

The dispersal of air pollutants is governed mainly by the mean wind speed and the characteristics of atmospheric turbulence. Together these factors determine the stability characteristics of the atmosphere. Turbulence consists of horizontal and vertical eddies that mix pollution with the surrounding air. As turbulence increases, so too does mixing, which results in good dispersion and lowered pollutant concentrations. Strong turbulence occurs in an unstable atmosphere in which vertical motion is enhanced. Conversely, when turbulence is weak, there is little propensity for mixing to occur and pollutant levels may increase. Conditions of weak turbulence are often typical of clear mornings in winter.

Turbulence can be generated through both thermal and mechanical means. Thermally generated turbulence results from heating of the surface, which leads to buoyancy-induced convective motion (i.e. vertical transport of heat away from the surface). With stronger heating of the surface, thermal turbulence will increase. Mechanically generated turbulence can result from wind shear as the wind blows across the Earth’s surface. Mechanical turbulence increases with increasing wind speed and is greater over rougher surfaces (e.g. cities and forests). For a more in-depth explanation of the theory and mathematical representation of turbulence in the boundary layer refer to Stull (1998) or Garratt (1992).

A typical diurnal cycle of radiative heating and cooling in cloud-free conditions results in varying stability conditions throughout the day (Figure A18). In the early morning, before sunrise, the nocturnal temperature inversion is still present. Any pollutants emitted close to the ground may be trapped in this inversion, if the chimney stacks are lower than the inversion top, smoke plumes will take on a fanning shape (Figure A18a). After sunrise, surface heating results in the generation of convection which mixes the air close to the surface within the layer beneath the remains of the nocturnal inversion. This situation is one of the worst for pollutant concentrations at ground level as fumigation occurs (downward movement of pollutants from aloft because of a growing mixed layer), resulting in pollution originally emitted and trapped aloft being mixed to ground level, often in quite high concentrations (Figure A18b). As daytime heating continues the nocturnal inversion is completely eroded and replaced by a lapse profile (decrease of temperature with height; Figure A18c). This is typical of an unstable atmosphere which promotes vertical mixing. This mixing is evident in the behavior of smoke plumes which may take on looping behavior (Figure A18c). In late afternoon the atmosphere may become neutral, in which the environmental lapse rate equals the dry adiabatic lapse rate (Figure A18d). This results in coning of smoke plumes (Figure A18d). As the sun sets the surface begins to cool and the nocturnal inversion is again established (Figure A18e). If the smoke stack is above the inversion height, then lofting of the plume is observed (Figure A18e).

A common and relatively simple estimate of stability is the Pasquil-Gifford stability classification (Table A13). This scheme assumes that stability in the layers near the ground depends on net radiation as an indication of convective eddies, and on wind speed as an indication of mechanical eddies. At low wind speeds and with strong daytime insolation (i.e. radiative heating from the sun), the atmosphere will be extremely unstable (class A). As both cloud and wind speed increase, the atmosphere becomes less unstable (B, C) and may become neutral (D). Under light winds and clear skies at night the atmosphere may become stable (E and F).

Figure A18

Stability changes throughout the day and effects on smoke plumes (modified after Whiteman, 2000). The dashed line (DALR) is the dry adiabatic lapse rate and the solid line is the environmental lapse rate.

Air pollution episodes are typically worse during stable atmospheric conditions since these restrict or limit the amount of vertical mixing. Temperature inversions, which are characteristic of a moderate to strongly stable atmosphere, can occur through several different mechanisms including surface cooling, warming of air aloft and advection (mean horizontal movement of air). Surface or radiation inversions occur during clear, calm nights as the ground cools rapidly due to the loss of long-wave radiation to the outer atmosphere. Their strength and depth may be locally increased by cold air drainage. Surface cooling can also occur through evaporative cooling as water evaporates from a moist surface during the day in fine weather.

Inversions due to warming can occur through subsidence during anticyclones (as described earlier) or in the lee of mountains. In both cases the descending air warms adiabatically and creates elevated inversions which can effectively cap any mixing of pollution from below. In the lee of the Rocky Mountains in winter, radiative cooling over the interior of North America creates a very cold air mass which ponds against the mountains. Descending warm air spreads over the top of this stagnant pool and severely inhibits any upward dispersion (Oke, 1992).

Advection inversions can occur when warm air blows across a cooler surface such as a cold land surface, water body or snow cover. In this situation the cooling of the underside of the air mass creates an inversion with its base at the surface (Figure A19a). Existing inversion structures can also be modified during advection. For example, when a cool sea breeze (or rural air) flows over warmer land (or urban) areas, the inversion becomes elevated, and this can lead to serious fumigation problems if chimney stacks are located in the coastal zone (Figure A19b, c).

Figure A19

Examples of advection inversions. (a) Warm air flowing over cooler water, (b) sea breeze fumigation on a clear day and (c) fumigation over an urban area (modified after Oke, 1992 and Karl et al., 2000).

Complex terrain such as in coastal or alpine regions, or indeed anywhere that has significant variation in topography, can lead to complex airflows. In such areas under clear skies and light synoptic winds, local circulation systems often develop, such as land-sea breezes or mountain-valley winds. These circulation systems are not good pollution ventilators since they tend to have low wind speeds, are often closed systems and have a diurnal reversal of flow. Thus the pollution can be recirculated over the source area and may build up to high levels.

Pollutant transformation

Air pollutants may often undergo chemical transformations in the air to produce what are termed secondary pollutants. The damaging London smogs of 1952, which killed thousands of people, involved sulfuric acid fogs which remained stagnant for 4–5 days. These fogs occur when SO₂ (which is generated by the combustion of fuels — particularly from coal with high sulfur content) is oxidized in the air to sulfur trioxide (SO₃) which then reacts with water vapor (H₂O) in the presence of catalysts to form sulfuric acid mist. If the meteorological conditions are such that dispersion is poor, these toxic mists can have deadly consequences.

More common in recent decades is the formation of photochemical smog. There is a naturally occurring NO₂ photolytic cycle in which solar radiation results in the photo-dissociation

Table A13 Pasquil-Gifford stability class

of NO₂ into NO and O (Figure A20). The oxygen atom (O) is highly reactive and combines with an oxygen molecule (O₂) to produce ozone (O₃). The ozone then reacts with NO to give NO₂ and O₂. However, the presence of reactive organic gases (ROG — which may be sourced from vehicles, but also naturally from vegetation) disrupts the cycle and can accelerate the production of ozone and other chemicals, such as aldehydes and ketones and peroxyacetyl nitrates (PAN) and even some particulates. Photochemical smog has a characteristic odor (partly due to the aldehydes), a brownish haze (due to NO₂ and light scattering by particulates) and may cause eye and throat irritation and plant damage. Figure A20 illustrates the sequence of chemical reactions that lead to formation of ozone as a by-product of photochemical smog following the emission of NO and ROG (the primary pollutants) into the atmosphere. Organic peroxy radicals (RO₂) are formed by the break-up of ROG, which by reaction with NO form ozone. This type of smog has been reported from many cities around the world, including cities in Australia, Brazil, Britain, Canada, Greece, Netherlands, Iran, Israel, Japan and the United States.

Figure A20

Chemical reactions in a polluted troposphere leading to photochemical ozone formation.

Pollutant removal

As well as through chemical reaction, pollutants can be removed from the atmosphere by other processes, such as gravitational settling, and dry and wet deposition (Figure A15). Gravitational settling removes larger particulates, with the rate of removal related to the size and density of the particles and the strength of the wind. This process removes most particulates larger than 1 micrometer (μm) in diameter, but smaller particles become influenced by turbulence and may remain aloft for long periods. Dry deposition is a turbulent process in which there is a downward flux of pollutants to the underlying surface. The amount of deposition depends on the characteristics of turbulence, with increased rates under stronger turbulence.

Wet deposition involves the removal of air pollutants through absorption by precipitation elements (water droplets, ice particles and snowflakes) and consequent deposition to the Earth’s surface during precipitation. This removal process includes the attachment of pollutants to cloud droplets during cloud formation, incloud scavenging (also known as washout), and through coalescence of rain drops with material below the cloud (rainout; Arya, 1999). Slin (1984) provides a comprehensive review of precipitation scavenging processes.

Receptor response

Air pollution is a concern due to its adverse effects on organisms and damage to materials. Research has demonstrated that air pollution can affect the health of humans and animals, damage plants, reduce visibility and solar radiation, and affect weather and global climate.

Effects on humans, animals and vegetation

During extreme air pollution episodes in urban areas the concentration of air pollutants can reach high levels for several hours or days. This can cause extreme discomfort, exacerbate illnesses and increase mortality rates among the most vulnerable part of the population (children, the elderly and the sick). Health-effect studies have revealed a number of adverse effects associated with common air pollutants. Peroxyacetyl nitrates (PAN), which are a component of photochemical smog, can cause eye and throat irritation. Other health hazards of air pollutants include chronic bronchitis, pulmonary emphysema, lung cancer and respiratory infections. Even short-term exposure to particulate matter can cause an increase in the rate of asthma attacks.

Adverse effects of air pollution on vegetation include leaf damage, stunting of growth, decrease in size and yield of fruit, and severe damage to flowers. Domesticated animals such as cattle and dogs cannot only breathe in toxic air, but may also ingest pollution-contaminated feeds. The air pollutants that are of major concern are fluoride, lead, and other heavy metals and particulate matter.

Effects on materials

Air pollutants affect materials by chemical reactions and soiling. Extensive damage can occur to structural metals, building stones, fabrics, rubber, leather, paper, and other materials. For example, a well-known effect of photochemically produced ozone in the troposphere is cracking of rubber products. The number of cracks in automobile tyres, as well as their depth, has been related to ambient ozone concentrations.

Air pollution dispersion models

During the past two decades advances in computer technology have allowed the use of increasingly sophisticated air pollution dispersion models (also known as air quality models). Numerous books are now available that give an in-depth explanation of the theory and derivation behind the mathematical formulations used in air pollution models. For example, Zannetti (1990) provides information on a wide spectrum of dispersion models, while Jacobson (2000) offers an excellent and exhaustive treatise on the theory, the numerical techniques and chemistry of air pollution models. Arya (1999) also describes the diffusion/dispersion theories behind the modeling systems including detailed mathematical formulation. These models are used for both regulatory and research purposes.

Box models

The simplest air pollution models are single- and multi-box models (Lettau, 1970). Box models predict the concentration of primary pollutants inside a box using mass conservation principles, where the box represents a large volume of air over a city. A box model considers the emission source at the surface (lower boundary), advective inflow and outflow at the sides (lateral boundaries), and the evolution of the mixing depth and subsequent entrainment of pollutants from aloft. More complicated box models include chemical and photochemical transformations to predict concentrations of secondary pollutants like ozone (Demerjian and Schere, 1979; Dodge, 1977).

Figure A21

Schematic of plume dispersion from a stack with Gaussian distribution in horizontal and vertical (H_s is the height of the plume centerline and hs is the stack height) (modified from Arya, 1999).

Gaussian plume models

Gaussian plume and puff models are used to estimate three-dimensional pollutant concentrations resulting from emissions generated at a point source, under stationary (unchanging) meteorological situations (Gifford and Hanna, 1973; Turner, 1964), although some models also incorporate non-stationary conditions. The underlying assumption in the Gaussian model is that with sufficiently long averaging periods (e.g. 1 hour) a bell-shaped (Gaussian) concentration distribution is produced in both horizontal and vertical directions (Figure A20). A plume model assumes a continuous point source with a uniform strong sbackground flow, where release time and sampling time are longer than travel time. Puff models are used in cases where emission from a source is almost instantaneous and much shorter than travel time of the plume, such as accidental releases of hazardous material into the atmosphere. Gaussian models are commonly used in a regulatory framework because of their simplicity and cheap computational demands (relative to three-dimensional numerical models, which are described below). The severe limitation of the Gaussian models is that they generally do not work in calm conditions; in addition, the complexity of local terrain and variation of atmospheric conditions with time is ignored.

Numerical meteorological models

The most sophisticated and hence computationally demanding way of calculating pollutant concentrations is by simulating detailed meteorology (three-dimensional fields of wind, turbulence, temperature and water content) and coupling these fields to emission scenarios. The role of the meteorological component of the model is to provide input for the pollutant diffusion component. A numerical meteorological model uses numerical techniques to approximate meteorological equations that describe fluid flow — collectively known as the Navier-Stokes equations. These are highly complex and nonlinear, and at present, no analytic solution is known for them. Stull (1998) and Pielke (2002) describe the numerical techniques and turbulence parameterizations behind these types of models. The Navier-Stokes equations take into account conservation of mass, heat, momentum and water. Air pollution scientists use many names for numerical models, but they are frequently referred to as grid-based and mesoscale.

Diagnostic models

Diagnostic models by definition do not provide a forecast for the state of the atmosphere. Diagnostic outputs are three-dimensional fields of meteorological variables obtained by interpolation and extrapolation of available meteorological measurements. They are computationally economical and appear to be effective analysis tools when the dominant factor driving the boundary layer airflow is the terrain. A major drawback is the requirement for sufficient observational data as input for the analysis. A three-dimensional diagnostic model such as CALMET — which is the meteorological component of the CALPUFF modeling system (Scire et al., 1990) — is widely used by regulatory/environmental organizations such as the US Environmental Protection Agency. The Earth Tech Corporation

Table A14 Summary of air quality standards from Canada, United States, New Zealand, and United Kingdom

freely distributes the code for the CALPUFF modeling system at http://www.calgrid.net/calpuff/calpuff1.htm.

Prognostic models

Prognostic models are able to forecast the evolution of the state of the atmosphere by numerically solving (both in time and space) the detailed set of approximations to the Navier-Stokes equations. Prognostic models take into account the evolution of synoptic scale winds and allow the formation of diurnally reversing local scale circulations such as sea/land breezes and valley/mountain flows (Whiteman, 2000). A great appeal of prognostic models, aside from their research applications, is that they can be the core of real-time air quality forecasting systems in case of accidental or intentional release of pollutants into the atmosphere. The forecast for the meteorological variables, in conjunction with appropriate turbulence statistics, is used to determine the probable spread and diffusion of the plume. Scientific centers that generate such forecasts and issue warnings use supercomputers, since expediency is important and prognostic models are computationally demanding. Examples of such organizations include the Atmospheric Research Division at the Commonwealth Scientific and Industrial Research Organization (CSIRO), which has implemented the Australian Air Quality Forecasting System (AAQFS), and the Lawrence Livermore National Laboratory in the United States which has the National Atmospheric Release Advisory Centre (NARAC). AAQFS routinely predicts daily levels of photochemical smog, atmospheric particles and haze, whereas NARAC is generally designed for emergency response purposes in case of accidental emissions of hazardous pollutants (e.g. radiological, chemical and biological substances).

Air quality guidelines and standards

Good health is (should be) a fundamental human right. The primary aim of air quality guidelines is to protect public health from the harmful effects of air pollution, and to eliminate or substantially reduce exposure to hazardous air pollutants. To achieve these aims, many countries use regulatory control through legislation to set legally enforceable air quality standards (AQS). These represent values that are the maximum average ground-level concentration allowed by law. AQS values are derived from air quality guidelines (AQG) determined purely from epidemiological/ toxicological analysis. AQS have to consider technological feasibility; cost of compliance; and social, economic and cultural conditions. The maximum concentrations allowed are categorized by averaging periods, the most commonly used being hourly, daily and yearly. The air quality standards are different in each country. For example, in the United States they are referred to as National Ambient Air Quality Standards (NAAQS), while in Canada they are known as National Air Quality Objectives. Currently, the maximum allowable concentration for each pollutant (and for each averaging period) is prescribed differently in each standard. Table A14 shows maximum allowable concentrations for some common pollutants.