View all publications

The State of New Zealand's Air

The air quality indicators which are commonly measured throughout the world have all been monitored in New Zealand at some time or other. Only a few have been monitored systematically, however, and these are summarised here.

Levels of air pollution are commonly reported in two different types of units which are not interchangeable because one measures volume while the other measures weight. They are:

  1. parts per billion (ppb) or parts per million (ppm)

    This is a volumetric measurement and refers to volumes of the pollutant (gas) per billion, or million, volumes of air. For example, a concentration of 50 ppm carbon monoxide is equivalent to 50 millilitres of the gas in 1 million millilitres (1 cubic metre) of air.
  2. micrograms per cubic metre (µg/m3) or milligrams per cubic metre (mg/m3)
    These are mass measurements and refer to the weight of pollutant (gas or particles) in 1 cubic metre of air.

    Most of the air quality measurements presented here were taken over the last 10 to 20 years. Where appropriate, the most recent data have been given greater emphasis. More detailed air quality data have been published by the Ministry for the Environ-ment (1994a) and Graham and Narsey (1994). Although Box 6.3 contains a brief discussion of indoor air quality, the substance of this chapter deals with outdoor air. The pollutants discussed are those which commonly present problems in ambient (outdoor) air, namely, particulate matter, sulphur dioxide, carbon monoxide, oxides of nitrogen, hydrocarbons and ozone, and lead.
Box 6.2: The hazards of lead-based paint dust

Lead is a heavy metal which is widespread in the environment. When absorbed into the human body, lead mimics calcium and tends to accumulate in the bones and teeth. Its half-life in the human skeleton (i.e. the time taken for its concentration to halve) is about 18 years (Atherley, 1978). At natural levels lead accumulation in the body seems to have no effect on human health. Comparisons of modern and prehistoric human bones reveal that our bodies now contain 5001,000 times more lead than the natural levels of our forebears (Patterson, 1982). Even at this level the lead burden seems to have no effect on us. However, at three or four times the current 'normal' level, lead becomes increasingly toxic. Early symptoms of lead poisoning include anaemia, loss of appetite, gastric pain and constipation. Since Roman times the signs of acute lead poisoning have been well known, beginning with the tell-tale blue or black 'leadline' around the gums (Nriagu, 1983). The Romans had very high lead levels because they ate and drank from lead plates and cups.

The poison affects the entire body, most notably, the brain (in severe cases causing coma, convulsions, blindness, deafness, and severe mental disorder); the gut (causing acute stomach ache or 'lead colic', loss of appetite, nausea, vomiting and constipation); the peripheral nerves (causing weakness of fingers and toes, drooping of wrists and feet, and ultimately paralysis or 'lead palsy'); and the blood (causing anaemia and possible kidney problems) (Atherley, 1978). At lower doses, elevated lead levels in children appear to be associated with impaired learning ability and behavioural disorders, such as hyperactivity and irritability (Fergusson, 1992; Hay and de Mora, 1993; Rutter and Russell-Jones, 1983; Tong et al., 1996).

Until the 1950s, a quarter of all brain-damaged lead poisoning victims died, and about half the survivors were mentally impaired. Today, early efforts are made to identify children who may be at risk from lead poisoning so that the hazards can be minimised or avoided before serious exposure occurs. Lead can be gradually removed from the tissues through lengthy treatment with substances containing calcium salts, but recovery is seldom complete. Doctors who diagnose blood lead levels above the statutory maximum values in an individual are required to notify the local Medical Officer of Health, but no statistics are kept in New Zealand on the incidence of lead exposure or lead poisoning. The average number of hospitalisations declined steadily from 14 per year in 1970-74 to 10 in 198589 (Public Health Commission, 1994). More than half the hospitalisations involved children aged one to four years. Most of the rest involved people in high risk environments (e.g. rifle ranges, battery factories) or handling high-risk products (e.g. lead pipes, lead-based paints).

Leaded petrol in cars has been the main source of airborne lead for most of us, but for people with lead poisoning, this 'normal' lead load has been topped up by additional lead sources. These include occupational sources, such as those already mentioned, and domestic sources, especially leaded paint flakes and dust in old houses. Children are particularly at risk of absorbing lead by inhaling dust or fumes (from the sanding or burning off of old paint) and by swallowing dust, paint flakes, or contaminated soil. Swallowing leaded surface dust (settled dust) is the most common cause of elevated lead concentrations in young children. Accessible dust contaminates the child's hands during play, and is transferred to the mouth via repetitive hand-to-mouth activities, such as thumb sucking and nail biting (Ministry of Health, 1996a).

Until 1965, many paints sold in New Zealand had high levels of lead. After that date, some lead-based paints continued to be sold, but in much smaller quantities. White lead was extensively used as a paint pigment until 1945 when it started to be phased out in favour of titanium dioxide. Nonetheless, white lead remained on sale in white undercoat until the mid-1960s. Lead sulphate was also the pigment used in pink primer until the mid-1960s. Lead chromate (yellow pigment) remained an ingredient in domestic paint until the late 1970s. Red lead steel primer is known to have been used as a wood primer until the 1980s. Calcium plumbate has been widely used as a coating for iron roofs from 1958 until the present day, but it is now no longer made, and limited stocks remain (Department of Labour, 1995).

Thus, although the concentration of lead in domestic paints has declined dramatically in the past 30 years, it may be assumed that pre-1970 interior or exterior domestic paintwork is almost certainly lead-based, whilst pre-1980 paintwork may be lead-based. An estimated 251,000 New Zealand houses may have lead paint on or in them, and each year some 5,000 of these undergo work to remove old paint (Jansen, 1984). Unfortunately, it is not possible to identify lead-based paint by its appearance, but it can be identified by simple spot tests. If sufficient care is not taken when removing it, people may be exposed to dangerous levels. 'Do-it-yourselfers' often overlook the dangers associated with preparing the family home for painting. Very young children and pets, especially dogs, run an even greater risk of contamination because they are more likely to ingest lead particles directly and in larger quantities. The Occupational Safety and Health Service (OSH) of the Department of Labour, and the Ministry of Health and public health units of Crown Health Enterprises (CHEs), are able to provide information on the health and environmental hazards linked to lead-based paint. The Ministry of Health (1996a) has issued guidelines for managing children's lead exposure.

Particulate Matter

The term 'particulate matter' refers to any airborne material in the form of particles, and encompasses pollutants we commonly refer to as dust, smoke, or aerosols. Particles less than 10 microns in size are generally known as PM10. PM10is an important health determinant because of its ability to penetrate deeply into the respiratory tract and lungs. Airborne particulate matter can arise from a wide variety of sources including domestic fires (especially coal and wood), power stations (coal or oil), motor vehicle emissions, rubbish burning, agricultural activities, quarries, road construction, building construction, and numerous industrial processes as well.

Natural sources of particulate matter include volcanoes (such as Mount Ruapehu's recent eruptions), sea spray, plant and animal matter (e.g. pollens and fungal spores), and wind-blown dust and dirt. High concentrations of airborne particles can have various effects, including reduced crop production, when large amounts of dust are deposited in rural areas.

To humans, airborne particles can be both a nuisance and a health hazard. The nuisance effects of particulate matter include the deposition of dust and grime on vehicles, buildings and other surfaces, and reduced visibility. Large quantities of particles can make the air look hazy or dirty. When combined with sulphur dioxide emissions, particulate matter can not only also cause soiling of building surfaces, but also corrode them.

The health hazard arises when wind-blown dust irritates the eyes, and when small particles are drawn deep into the respiratory tract. Long-term exposure can seriously impair the performance of the lungs. Those most at risk are people with chronic lung diseases, asthmatics, the elderly, and very young children. In fact, an increasing number of overseas studies have shown that the fine particulate matter which comes from vehicle exhausts causes increased death rates in people with lung and heart ailments (D. Bates, 1996; Anderson et al., 1996; Dockery and Pope, 1994; Health Effects Institute, 1995; Department of Health, 1995). One British estimate put the number of increased deaths in England and Wales at 10,000 per year (Bown, 1994a; Hamer, 1995). The strength of the medical evidence has persuaded the British Department of the Environment to recommend a new air quality standard for PM10of 50 micrograms per cubic metre (50 µg/m3) for a 24-hour period (Department of the Environment, 1995).

The current New Zealand guideline for 24 hours is 120 µg/m3(Ministry for the Environment, 1994a). The typical background level for rural areas in New Zealand is less than 1 µg/m3and for an urban neighbourhood 25-30 µg/m3. Readings of more than 300 µg/m3, well in excess of the New Zealand guideline, are comparable with cities in other parts of the world (Fisher and Thompson, 1996).

Airborne particulate matter can be measured in different ways to reflect some of the effects noted above. The most commonly measured indicators of particulate pollution are:

  • dust deposition
  • suspended particulate
    • total suspended particulate(TSP)
    • inhalable particulate (PM10)
  • smoke.

Dust Deposition

Dust deposition is the amount of dust settling over a fixed surface. It was monitored at more than 40 sites around the country during the 1970s at a variety of locations including Auckland, Waiuku, Meremere, Paeroa, Kawerau, Karamea and Christchurch.

Most of these measurements were directed at specific industrial sites, and most had been discontinued by the end of the decade. The only significant monitoring in recent times has been carried out by various mining companies in areas such as Huntly, Waihi and Central Otago.

Complaints about dust fallout generally occur when deposition levels exceed 4 grams per square metre over 30 days (4 g/m2/30 days). The Ministry for the Environment's Ambient Air Quality Guidelines do not set a guideline for dust deposition because it is regarded as more of a nuisance than a health issue. At what point a nuisance becomes an environmental issue is a matter for local authorities and their communities to judge.

Monitoring has shown that the levels of dust deposition vary widely. It is difficult to interpret some of the more extreme results without detailed knowledge of the individual monitoring programmes. However, some general observations can be made from the data:

  • Background deposition levels in relatively clean environments are generally less than about 1 gram per square metre over 30 days (1 g/m2/30 days).
  • Deposition levels in urban areas not affected by specific industries, are typically around 1 to 3 grams per square metre over 30 days (1-3 g/m2/ 30 days).

Suspended Particulate

Total suspended particulate (TSP) refers to particles floating in the air. It is measured by drawing an air sample through a filter and weighing the amount of particulate collected.

Inhalable particulate (PM10) which is the measure preferred by public health authorities, is similar to TSP but only refers to particles with a median diameter of 10 microns or less which can be absorbed into the lungs.

Suspended particulate matter has been measured at more than 80 sites throughout the country over the last 30 years. Locations include Whangarei, Auckland, Waiuku, Waihi, Huntly, Hamilton, Kawerau, New Plymouth, Hutt Valley, Wellington, Christchurch and Dunedin.

Measuring for more than 2 years has occurred at only about one-quarter of these sites. About half the sites could be considered as area monitors, with the remainder directed at specific industries. Most of the latter are still being maintained by the companies concerned.

The equipment used for most of these measurements was designed in New Zealand, and the results cannot easily be compared with measurements from many other countries. The New Zealand equipment gave results 25-50 percent lower than the standard equipment (high-volume samplers) used overseas (Graham and Narsey, 1994). Similarly, the results should not be compared directly with the Ambient Air Quality Guidelines for New Zealand. The guideline pollutant is PM10, whereas the New Zealand monitors measured TSP.

Most measurements taken using the local equipment have been evaluated against an unofficial standard of 60 micrograms of TSP per cubic metre (60 µg/m3) averaged over 7 days. More conventional monitoring methods will have to be adopted in future, however.

Some typical results for suspended particulate measured under various situations in New Zealand are summarised below, using 7-day averages. All of the data are for total suspended particulate:

  • Background levels of suspended particulate are typically below the unofficial standard, falling in the range of 20 to 50 µg/m3. Urban areas unaffected by any specific sources fall within the same range. (A notable exception to this is Christchurch, where wintertime levels are typically in the range of 40 to 80 µg/m3, with occasional excursions above 100 µg/m3).
  • Urban TSP levels in areas of significant industrial or commercial activity are typically in the range of 30 to 60 µg/m3, although occasional excursions up to about 100 µg/m3can occur.

Levels of suspended particulate matter have shown significant improvements over the last 2 to 3 decades at locations throughout the country. This applies both to general urban areas, and in the vicinity of major industrial sites. For example, at the Mount Albert (residential) and Penrose (industrial) sites in Auckland, the levels are now around half of their 1970 levels. Some of these changes are illustrated in Figure 6.5.

Figure 6.5: Total suspended particulate matter (annual averages) measured in the Auckland suburbs of Mount Albert (residential) and Penrose (industrial), 1964-95

Average annual total suspended particulate matter between 1964 and 1995 measured in an Auckland residential area (Mount Albert) is compared with that measured in an industrial area (Penrose). The industrial area has a higher concentration than the residential area, but concentrations in both areas have decreased by around half over that time period. However, there appears to be a slight increase in total suspended particulate at the Penrose site between 1992 and 1995.

Source: Institute of Environmental Science and Research Ltd.

Box 6.3: The State of our Indoor Air

Although air pollution is generally perceived as an ambient (outdoor) issue, indoor air quality can also cause serious problems and increase vulnerability to outdoor air pollution. A wide range of indoor environments exist. Workplaces and households are where most people spend most of their time, and the most common air contaminant in these environments is probably smoke from cigarettes, and also from malfunctioning gas appliances, wood burners and open fires. There is clear evidence that coal and wood fires increase the risk of upper and lower respiratory tract infections (Larson and Koenig, 1994). However, it is also known that children in homes with coal and wood fires are less likely to develop asthma and hay fever (von Mutius et al., 1996).

Alone, or in combination, cigarette, coal and wood smoke contain many toxic substances which increase the risk of cancers, bronchitis, emphysema, and heart attacks. Among the substances are particulate matter, dioxins, and carbon monoxide. Cigarette smoke also contains benzene, which is associated with acute non-lymphocytic leukaemia in children and adults. No studies have been done on the levels or concentrations of tobacco smoke in work and home environments. However, the Ministry of Health found in 1996 that nearly 20 percent of workers were exposed to tobacco smoke in their workplace, with approximately 35 percent of those exposed during tea breaks. The exposure rate in homes may be higher, particularly for children, because 23 percent of the adult population were regular tobbaco smokers in 1996. One-quarter of households with children aged under 5 years contained regular smokers. Nearly 85 percent of non-smoking adults live in households where nobody smokes. Thirty-two percent of smokers do not smoke in their own home. The smoking rate for the Māori population (38 percent) is much higher than for the Pacific Island population (22 percent) and Europeans and others (19 percent) (Ministry of Health, 1996b).

The Smoke-free Environments Act, which took effect in 1990, bans smoking in those indoor workplaces to which the public has access, and in most office areas, and encourages a smoke-free working environment. It bans smoking on public transport, and requires that at least half the space in restaurants must be 'non-smoking'. Trends in tobacco sales and public survey responses show that tobacco consumption is declining, though only a small number of smokers are actually giving up the habit. Between 1976 and 1993 New Zealand's consumption of tobacco declined faster than any other OECD country.

Smoking rates fell from 34 percent of the population to 27 percent. The decline was confined almost entirely to the European population, with Māori and Pacific Island smokers showing little change (Public Health Commission, 1994). The trend continued in 1994 and 1995, with total tobacco consumption declining a further 6.5 percent in the year, but few people quitting. The average smoker had 102 cigarettes a week, compared to 107 in 1995. The 1996 Ministry of Health survey shows that 21 percent of New Zealanders are smokers.

Smoke is not the only indoor air pollutant, however. Emissions from gas appliances and kerosene heaters may also affect respiratory health. Nitrogen dioxide, for example, a by-product of combustion (see Box 6.5), has been suspected of worsening asthma. Although British studies have found no link to asthma in humans (Bown, 1993), recent US research has found that very high levels of nitrogen dioxide (about ten times the normal human exposure) does aggravate asthma in rats (Kaiser, 1996). A recent survey of 40 New Zealand homes with unflued gas appliances found nitrogen dioxide levels close to or above the World Health Organisation (WHO) guideline of a one-hour average of 160 parts per billion (ppb) (Bettany et al., 1993). Levels were higher for homes with wall convective gas heaters than for those with portable conductive or portable radiant gas heaters or gas cookers. Homes with flued gas appliances, or none at all, had nitrogen dioxide levels similar to ambient air.

All homes and some school classrooms with unflued gas appliances sometimes exceeded the New Zealand indoor standard for carbon dioxide, indicating inadequate fresh air supply. There were sporadic instances of high carbon monoxide. On two occasions this was associated with flued gas appliances, reflecting either poor maintenance or faulty flue installation. Six of the 40 houses had high levels of formaldehyde, in two cases above the WHO level of concern of 0.1 ppm. Indoor sources of formaldehyde include composite wood products, furnishing fabrics, carpets, and gas appliances. Five of these six houses were less than 6 years old, and one had had recent renovations (Bettany et al., 1993).

Other indoor environments which are likely to have some air problems are industrial worksites where dust, fumes or odour are generated as part of the production process. Because workplaces are legally subject to occupational health and safety requirements, air problems in these locations are dealt with by employers and employees on a case by case basis. The overall extent of workplace air problems is unknown.

Box 6.4: Perceived Air Quality

Although the unaided eye or nose can sometimes detect polluted air, this is not an infallible method for assessing air pollution. Lead, carbon monoxide, and many other pollutants are invisible and odourless. Special instruments are therefore required to measure them. Instruments are also needed to determine the seriousness of a pollution episode. Sometimes air may look quite clean when, in fact, a serious problem exists. At other times it may look polluted when the levels of harmful contaminants are low.

In Auckland, for example, people often comment on the polluted appearance of the air above the city even when objective measurements sometimes reveal no problems. There are several reasons why perceived air quality may seem worse than measured air quality. For one thing, pollution is most likely to be seen during relatively short peak periods, while measurements are generally averaged over periods of 1, 8, or 24 hours. Also, measurements are often taken at single sites at ground level, while observed pollution is often well above the ground and spread over several kilometres.

The most commonly perceived air problems, however, are those detected by the nose rather than eye. In a country as sparsely populated and as windy as New Zealand, complaints about offensive odours should be quite rare. Such is not the case, and of all the issues relating to air quality, few are more likely to raise public interest-and anger-than strange and offensive smells. The Auckland Regional Council estimates that approximately 80 percent of all air pollution complaints it receives relate directly to odour (Allen, 1993). Other councils report a similar experience. The problem is not confined to the city. The bad smells come mainly from sites such as piggeries, mushroom farms, meat works, and wastewater. They often occur sporadically, depending on weather conditions and wind direction, and are difficult to assess. Odour from agrochemical sprays is also a common source of complaints in some rural districts.

The Ministry for the Environment has released a discussion document and a practice guide on odour measurement and management (Ministry for the Environment, 1994b and 1995). However, monitoring the problem is notoriously difficult because smell is a subjective sensation which cannot be objectively measured by instruments other than the human nose. Although it is possible to measure some odours from concentrated emission sources (e.g. smokestacks), the quality assurance procedures needed to get reproducible measurements are very difficult to implement.

Some councils have recorded the number and form of complaints about offensive smells (and noise), raising the possibility that complaints could be considered as an environmental indicator. However, complaint-based indicators are very subjective. An indicator of odour nuisance which is not based on complaints has been developed in the Netherlands where questions about the extent of odour nuisance are included in national social surveys dealing with a wide range of issues. By not placing undue emphasis on offensive smells, this provides a more balanced measure of the true extent and distribution of the problem.

Public perceptions of air quality problems are not always confined to things that can be seen or smelt. Extremely low frequency (ELF), or radio frequency (RF), electromagnetic radiation is invisible and odourless but public concern has grown in recent decades about whether it can cause cancer or other health problems in humans (Parliamentary Commissioner for the Environment, 1996). RF radiation is emitted by various sources, ranging from power lines and radio and television transmitters through to a whole array of household electrical appliances, such as microwave ovens, electric kettles, cellphones, computer screens and TV receivers.

RF radiation travels in waves which radiate out from the electric current like the ripples generated by a pebble dropped into a pond. The space the waves occupy as they radiate out is referred to as an electromagnetic field. The amplitude, or strength, of the field is determined by the strength of the electric current, but it weakens rapidly as it spread out from the source. The RF radiation in homes is mostly leakage from low-powered appliances (a TV, for example, uses 300 watts). However, large radio and TV transmitters are designed to carry signals over large distances and therefore have strong RF electromagnetic fields.

Cellphone transmitters use low levels of power (normally between 20 and 150 watts, or roughly light bulb strength). However, the increasing number of cellphone transmitters being erected in residential areas, and the uncertainty over whether there is a causal link between RF radiation and health problems, has stimulated public concern. For example, an application by a mobile phone company to erect a cellphone tower on a building next door to a Christchurch primary school was challenged in court. The case was lost, but it served to highlight the public's perception of the problem.

Scientists have yet to find clear evidence that RF radiation is harmful. A number of overseas studies have failed to agree on the effects, with some finding no effects at all. As scientists debate the issue, the World Health Organisation is recommending that high priority be given to further research.


The darkness of particles collected on a filter is measured using light reflectance. This gives an indication of relative 'soiling potential' and was originally used for monitoring smoke from domestic fires. Measurements are highly dependent on the physical properties of the particles collected (especially size and colour or darkness).

The results of the measurements are converted to micrograms per cubic metre (µg/m3) by means of a standard calibration curve, but this was originally derived from the United Kingdom and is really only applicable for communities where coal smoke from domestic fires is the predominant source.

The general practice in New Zealand has been to use this calibrating curve to convert the measured light reflectance into an estimate of smoke particle mass, but the results are expressed as 'smoke units' rather than micrograms per cubic metre. (One 'smoke unit' is equal to about 3 micrograms per cubic metre (3µg/m3), with minor variations according to season, location, and fuel type).

Smoke levels have been measured at more than 60 sites over the last 20 or so years, although many of these were short-term projects, operating for periods of no more than about 6 months to 2 years. About two-thirds of all the sites were in Christchurch, with the remainder in areas such as Whangarei, Auckland, Huntly, Dunedin and Invercargill.

As with suspended particulate matter, smoke levels around the country have also shown some improvements over the last 10 to 20 years. In Christchurch and Dunedin, for example, wintertime levels of smoke have decreased, significantly in the case of Christchurch, especially over the last decade. These improvements are shown in Figure 6.6.

The results from the more recent monitoring programmes can be summarised as follows:

  • Summer smoke levels throughout the country are typically less than 5 to 10 smoke units.
  • Current wintertime smoke levels in areas other than Christchurch show daily maxima typically in the range of 30 to 50 smoke units. (By comparison, wintertime smoke levels in Christchurch over the period 1980-85, showed daily maxima of up to about 200 smoke units.) In recent years, this type of monitoring has been upgraded to measure TSP and PM10, providing more useful information about health effects than smoke units do.

The above data cannot be compared directly with any air quality guidelines because of the nature of the measurements. Levels greater than about 30 smoke units would be visually assessed by most people as 'smoky' conditions, and results above about 100 smoke units would probably be cause for widespread concern. It is interesting to speculate on the reasons for this improvement in air quality. Quite clearly, if the change has occurred throughout the country in both residential and industrial locations, then it is probably not due solely to improvements in the control of individual pollution sources.

The most likely explanation for the improve-ments are changes in fuel use. Coal consumption has declined and use of electricity has increased. A rise in air pollution complaints to the Christchurch City Council in 1992 may have resulted from the greater use of domestic fires during the '1992 Hydro Power Crisis' (Brieseman et al., 1992). [The increase in sulphur dioxide emissions during the 'Power Crisis' would seem to bear this out (see Figure 6.8).]

The Clean Air Act came into force in 1972 but this was primarily concerned with the control of industrial emissions. Although there have been controls on smoke emissions from domestic fires in Christchurch, this has not been the case elsewhere. Smoke is still the main wintertime air contaminant in other Canterbury urban areas (Canterbury Regional Council, 1993).

In Wairarapa, emissions from domestic fires can have greater impact on overall air quality than other sources, such as industry. The population density is low, but an estimated 80 percent of Wairarapa homes have some form of solid fuel heating. Climatic conditions can cause the emissions, especially smoke, to accumulate and reduce visibility (Gazely and Bird, 1993).

Figure 6.6: Smoke levels (annual averages) for Christchurch (Manchester Street), 1973-87, Dunedin (corner Princes and Rattray Streets), 1975-88

Average smoke levels in both Christchurch and Dunedin decreased over the 1970s and 1980s. Throughout, Christchurch levels are markedly higher, and in 1987 were around 4 times higher than the levels in Dunedin.

Source: Institute of Environmental Science and Research

Sulphur Dioxide

Many air pollutants belong to a group called the oxides, which are formed when oxygen combines with other substances (see Box 6.5).

Sulphur dioxide (SO2) is an acidic gas which has a pungent odour in high concentrations. It arises from the burning of sulphur in coal and oil. Other carbon fuels, such as natural gas, petrol and wood, have insignificant amounts of sulphur.

The primary sources of sulphur dioxide are coal (typically between 0.5 percent to 3.0 percent sulphur, depending on type), fuel oil (ranging from 0.5 percent sulphur for light refined products, up to 4 percent for some heavy industrial types used in ships, power stations, and refineries), and diesel (0.3 percent).

Sulphur dioxide can also be emitted from a number of specific industrial operations, such as fertiliser and sulphuric acid manufacturing plants and oil refineries. It is also found in volcanic gases.

Sulphur dioxide is an irritant which can affect breathing and possibly harm the respiratory system. High concentrations can cause cell damage to plants. When combined with moisture in the air, sulphur dioxide forms sulphuric acid, a corrosive which can damage building and other materials.

Prolonged exposures to mixtures of sulphur dioxide and inhalable particulate matter may be linked to increased cases of respiratory diseases such as bronchitis, particularly in young children. When present in the air, there is usually particulate matter present as well. As a result, the effects of these two pollutants are difficult to separate, and their combined effects are worse than the effects of each individually.

Sulphur dioxide has been measured at more than 40 sites over the last 20 or so years, although about half of these were only for short periods. Much of the effort has been in Christchurch, with the remainder in Whangarei, Auckland, Huntly, Dunedin and Invercargill.

Prior to 1987, most monitoring was carried out using a wet-chemical procedure which gives a measurement of 'total acidity' as a 24-hour average result. Instrumental monitoring for sulphur dioxide has gradually taken over as the preferred method since that time in Auckland, Huntly, Christchurch, Dunedin, and Invercargill.

Monitoring in Auckland has mainly been based around 3 sites. These indicate that sulphur dioxide levels have declined significantly over the last 10 years or so. Prior to about 1980, sulphur dioxide levels averaged about 15 to 20 micrograms per cubic metre, (15-20 µg/m3) with daily maxima of up to 75 µg/m3. Levels in more recent years have been around 5 µg/m3or less, with daily maxima of up to 20 to 30 µg/m3.

In Huntly, sulphur dioxide is monitored at 3 sites, with 2 of these operating since 1977. Once again, the levels have declined significantly over the last decade, with initial average levels of 10 to 15 µg/m3now dropping to 5 to 10 µg/m3, and the daily maxima of up to 75 µg/m3now down to 15 to 20 µg/m3. This monitoring is carried out by the Electricity Corporation of New Zealand (ECNZ).

In Christchurch, a total of 20 or more sites have been used for sulphur dioxide monitoring, although about two-thirds of these were only operated as short-term sites in the mid-1970s and usually over the winter months of each year. Since 1987, monitoring has been conducted at only one instrumental site, a GEMS/AIR site in Packe Street.

As with the other cities that have been monitored, Christchurch's sulphur dioxide levels have improved markedly over the last 10 to 20 years. Averaged across the year, sulphur dioxide levels in the mid 1980s were in the order of 15 to 30 µg/m3, with maximum daily values in the range of 50 to 100 µg/m3. Prior to about 1980, most Christchurch sites were recording results of about double these values. Since 1987, the suburban Packe Street street in Christchurch has shown annual averages of 3 to 7 µg/m3, and maximum daily values of up to 30 µg/m3. However, care is needed in comparing these results with the earlier ones because of the differences in methodology and location.

In Dunedin, monitoring for sulphur dioxide has been conducted mainly at two sites, although an additional six have been used for short-term studies. The results indicate annual average levels of 10 to 20 µg/m3with maximum daily values of 35 to 70 µg/m3. Once again, these values are for data from the mid- to late 1980s, which are significantly lower than those for the previous decade.

In Invercargill, monitoring at two sites in the mid-1980s indicated relatively low levels of sulphur dioxide, with annual averages of 5 to 10 µg/m3and daily maxima of no more than 25 µg/m3.

Sulphur dioxide levels have dropped significantly over the last two decades, and, when compared to the New Zealand Ambient Air Quality Guidelines, levels are generally very low in most urban areas. This is clearly illustrated in Figure 6.7 showing average annual levels for Auckland, Christchurch, and Dunedin.

It should be noted, however, that there are situations where sulphur dioxide emissions from specific local sources have caused concern in recent years. The area around Marsden Point in Whangarei where there is an oil refinery is perhaps the most widely publicised of these.

A 1994 survey also showed that in Greymouth, on still winter nights when coal with a high sulphur content is burned in domestic fires, sulphur dioxide levels exceed the 24-hour recommended levels (West Coast Regional Council, 1994).

Figure 6.7: Sulphur dioxide levels (annual averages) for Auckland (Penrose) 1975-76 and 1978-96, Christchurch (Manchester Street) 1976 and 1976-87, and Dunedin (corner Princes and Rattray Streets), 1975-88

Average annual sulphur dioxide levels in Christchurch, Auckland and Dunedin have decreased over time. Christchurch has had the highest levels, and these almost halved between 1976 and 1988. Auckland's levels peaked in 1980 but have since more than halved and became lower than Dunedin 1982. The maximum acceptable level is 50 micrograms per cubic metre. In 1986, Christchurch's level was around 28 micrograms, Dunedin around 7 micrograms and Auckland around 5 micrograms.

Source: Institute of Environmental Science and Research Ltd.

The Greymouth survey shows that 2 sites were monitored, one for 97 days during which time the guideline level was exceeded 5 times (5.2 percent of time) and the other for 90 days during which time the guideline level was also exceeded 5 times (5.6 percent of time). (Although the United States has a less rigid guideline than New Zealand, exceeding the guideline more than once in a year would mean 'non attainment' and necessitate an extensive programme to clean up air quality within three years.)

During the 1992 winter power crisis when very low levels in the South Island hydro scheme lakes forced the Electricity Corporation of New Zealand to burn more coal and oil to maintain power supplies, there was a sharp jump in sulphur dioxide emissions (see Figure 6.8).

Figure 6.8: Estimated sulphur dioxide emissions resulting from oil and coal-fired electricity generation, 1980-96

Estimated sulphur dioxide emissions from oil and coal-fired electricity generation have fluctuated between 1980 and 1996. Overall, the proportion of sulphur dioxide emissions coming from coal has been greater than from oil. Although the proportions were closer during the 1982 and 1982 peaks. The peak in 1992 was the largest with emissions of nearly 1,800 tonnes resulting from the hydro-power crisis. In 1995 levels were around 800 tonnes compared to around 450 tonnes in 1980.

Source: Ministry of Commerce (1995), Baines (1993)

Box 6.5: How Oxides are Formed

Oxidation is a simple but vital process that happens all the time. It occurs when iron rusts and organic matter rots. It happens when fires burn, and even when we breathe. It also happens quietly in the air around us. Basically, it occurs whenever an oxygen atom combines with another atom (such as carbon or nitrogen or sulphur or iron) to form a new molecule. The new molecule is called an oxide. Often oxygen creates oxides by tearing other atoms away from a pre-existing molecule, causing the latter to burn or rust or rot. When it happens to the food inside us, we call the process respiration. Carbon is torn away from food molecules by the oxygen we breathe and exhaled back into the air as carbon dioxide. The energy and remaining atoms released by the departing carbon can then be absorbed by our bodies.

Some oxides are formed by only one oxygen atom, e.g. carbon monoxide (CO) and nitric oxide (NO), while others are formed by two of them, e.g. carbon dioxide (CO2), nitrogen dioxide (NO2), and sulphur dioxide (SO2). The shorthand terms for oxides of nitrogen and sulphur are NOxand SOx. In our cars, homes, factories, and in power stations using oil or coal, oxidation occurs through combustion. When fuel or waste containing carbon is burnt, carbon monoxide (CO) and carbon dioxide (CO2) are formed. Their relative amounts depend on the efficiency of combustion. More efficient processes create more CO2than CO. If trace levels of inorganic elements, like sulphur, are present in the fuel or air, these are also oxidised to produce sulphur dioxide (SO2). Because air is 79 percent nitrogen, this also oxidises during combustion to produce oxides of nitrogen (NOx). The newly-formed carbon dioxide and other oxides escape as fumes. Carbon monoxide is particularly dangerous because it is invisible, odourless, and toxic. With the rise of industrial society, and the heavy use of coal, oil, and petroleum, oxidation through combustion has increased dramatically and, with it, the levels of carbon, sulphur, and nitrogen oxides.

Carbon Monoxide

Carbon monoxide (CO) is a colourless, odourless, highly toxic gas, which is formed by the incomplete combustion of fossil fuels. When inhaled, it binds to haemoglobin in the blood, displacing oxygen. Prolonged exposure at moderate levels can lead to symptoms such as headaches and dizziness, while at high levels it can lead to loss of consciousness and even death.

At the lower levels which are typically encountered in urban areas it can serve as a useful indicator of the influence of vehicle exhaust emissions on air quality. Vehicles are the main source of carbon monoxide in most parts of New Zealand. Carbon monoxide can also be present in the emissions from domestic open fires and backyard incinerators, industrial fuel use and some specific industrial processes, such as steel manufacture.

In the past, carbon monoxide was monitored on a fairly sporadic basis, using short-term studies over periods ranging from several weeks up to about 6 months. Permanent monitoring sites have only been established in Auckland and Christchurch over the last few years.

Table 6.2: Non-volcanic sulphur depositions in New Zealand and other countries (grams of sulphur per square metre)
Location Average deposition (gS/m2/yr) Maximum deposition (gS/m2/yr)
Germany 6.5 >10
England 4.3 7
Norway 0.9 3
Sweden 1.2 4
Australia (four states only) 0.2 ?
New Zealand 0.15 (model) 0.7 (estimated)

Source: Holden and Clarkson (1985)

The first significant studies of carbon monoxide were carried out in Auckland and Christchurch in the early to mid-1970s. These indicated that this pollutant could accumulate to undesirable levels for significant periods of time in some inner city streets.

Measurements inside a shop in Queen Street (the main street through Auckland's central business district) during 1974 showed that the World Health Organisation (WHO) 8-hour guideline of 10 milligrams per cubic metre (mg/m3) was exceeded about 35 percent of the time (World Health Organisation, 1987). Similar results were reported for a short-term monitoring exercise in Hamilton during 1978.

Only a limited amount of carbon monoxide monitoring was carried out in Auckland and Hamilton during the 1980s, but this indicated that the situation was probably much the same as before. A new site was established in Auckland in 1990, again located in Queen Street, but well away from major intersections.

The levels of carbon monoxide now being recorded are much lower than in the past, partly due to the new site and partly due to the fact that traffic flows in and near Queen Street are now lower than those in the 1970s. During 1993-94, however, the guideline was exceeded 4 times at the new Queen Street site.

The regional council placed a second carbon monoxide monitor on Dominion Road in April 1994. During the 8 months it operated, the WHO guideline of 30 mg/m3 for 1 hour, or 10 mg/m3 over an 8-hour period, was exceeded 16 times. Traffic flows on this road are roughly 3 times the volume on Queen Street.

Carbon monoxide monitoring was also carried out in Christchurch during the mid-1980s. However, this was at a suburban site, rather than in the inner city area, and the carbon monoxide levels were relatively low.

The new monitoring site, in the suburb of St Albans, is also away from the city centre. Even so, the WHO guideline has been exceeded between 4 and 17 days per year at this site since 1988, generally on winter nights, suggesting that domestic fires may be partly responsible. During 1993, the regional council monitored carbon monoxide along Riccarton Road, a location with high traffic flows, for 61 days. The guideline was exceeded 56 times during this period (see Figure 6.9).

Figure 6.9: Carbon monoxide (CO) levels (static hourly average) in Riccarton Road, Christchurch, showing typical variations for traffic movement on days of low wind speed.

Carbon monoxide levels change at a similar rate to the number of vehicles on the road on days of low wind speed, both peaking during the morning and evening rush hours. Between about 4.30pm and 6.30pm the carbon monoxide level is above the acceptable 1 hour average of 30 milligrams per cubic metre, and from around 8am to 11pm the level is above the acceptable 8 hour average of 10 milligrams per cubic metre.

Source: Foster (1994)

At present there is no reason to believe that carbon monoxide levels near busy inner city intersections in New Zealand will be any lower than those recorded in the past. Increases in vehicle numbers are more likely to raise levels than to lower them.

This is borne out by some recent measurements in Gisborne and Whangarei which showed that carbon monoxide levels in the central business district could be moderately high even in cities of this size (Gisborne 31,000; Whangarei, 44,000) (Gisborne District Council, 1993; Northland Regional Council, 1994).

Monitoring in Gisborne showed that for 25 percent of the time over 20 sample days, carbon monoxide levels were unhealthy. In Whangarei, high levels of carbon monoxide occurred at one of two sites monitored for 9 percent of the time over 90 sample days.

Recent studies in Auckland and Christchurch have found that carbon monoxide levels in busy traffic corridors can exceed the ambient air guidelines in both summer and winter, whenever wind speeds are low and traffic density is high (e.g. Foster, 1994). Motor vehicles are the dominant source. However, ambient air pollution away from the traffic corridors occurs only in winter when emissions from household fires combine with those from motor vehicles. In Christchurch, for example, motor vehicles and domestic fires contribute about equally to ambient wintertime carbon monoxide pollution.

Figure 6.10: Sample measurements of nitric oxide (NO), nitrogen dioxide (NO2), and ozone (O3) from Christchurch, 6-7 July 1996

Levels of nitric oxide, nitrogen dioxide and ozone fluctuate throughout the day, occasionally they breach the recommended in guideline values to protect health.

Source: NIWA (1996)

Box 6.6: Acid Rain-A Northern Hemisphere Problem

Acid rain, which is a serious problem in some Northern Hemisphere countries, occurs when oxides of sulphur and nitrogen, which arise from the burning of fossil fuels, react with moisture in the atmosphere to produce a corrosive rain. Its worst effects are seen in Europe, Scandinavia and North America, where lakes and forests have been poisoned.

The acidity of rain is measured on the pH (powers of Hydrogen) scale which ranges from 0 to 14. Pure water has a pH value of 7, which is neutral. Values less than 7 are acid, and the stronger the acid the lower the pH. Orange juice has a pH of 4.5, vinegar has a pH of 3, and battery acid has a pH of 1. Values greater than pH 7 are alkaline. Rainfall is usually slightly acid with a natural pH of around 5.6. Acidification is only considered serious when the pH falls below 5. Rainfall with pH values of 5 to as low as 3 has been recorded in Europe and North America. In a comparatively unpolluted environment like New Zealand's, the pH of rain typically ranges between 5 and 6.

Only three sources of acid rain can affect New Zealand: the locally-produced sulphur dioxide from domestic and industrial sources; wind-blown imports from Australia; and the occasional belching of the North Island's volcanoes (e.g. Ruapehu). Holden and Clarkson (1985) estimated that, in an average non-volcanic year, the total fall-out of sulphur in New Zealand amounts to 42,000 tonnes, two-thirds of which blows over from Australia. This means an average loading of 0.15 grams of sulphur per square metre per year over the whole country-barely 13 percent of Sweden's level and less than 3 percent of Germany's (see Table 6.2). Holden and Clarkson concluded that acid rain is not a significant problem in New Zealand and is unlikely to become one at current levels of industrialisation and fuel combustion. In light of this, rainfall chemistry monitoring by the National Institute of Water and Atmosphere Research (NIWA) was terminated in favour of other research priorities.

However, on the rare occasions that significant volcanic activity does occur, the question of acid rain is inevitably revived. Until 1995, Mount Ruapehu had been relatively inactive for half a century. In September and October of that year it erupted, and did so several times in 1996 as well. In just two weeks in October 1995, the mountain spewed sulphur dioxide into the air at a rate of between 1,900 and 17,000 tonnes per day-three years' worth in a fortnight (Bell, 1995). Two vast ash falls sprinkled to the ground as far as Gisborne and Waipawa and a large sulphurous cloud stretched from Hawke's Bay down over the Wairarapa and Wellington, and out to sea between Christchurch and the Chatham Islands.

Vulcanologists did not expect any significant harm to the environment but predicted that vegetation away from the mountain could be affected if acid rain fell over a long period or gas emissions increased drastically (Bell, 1995). They cited the example of Hawaii where crops and natural vegetation were damaged over nine years of repeated eruptions. Soil scientists who analysed the ash falls concluded that they had useful amounts of sulphur and magnesium for fertiliser, "but heavy falls may have nuisance levels of sulphur and high soil acidity associated with them." (Cronin et al., 1996).

Oxides of Nitrogen

Oxides of nitrogen (NOx) is the term used to describe a mixture of two gases, nitric oxide (NO) and nitrogen dioxide (NO2). These are formed in most combustion processes by the oxidation of the nitrogen present in the air (see Box 6.5). Nitric oxide is the primary product but this can then be further oxidised in the ambient air to form nitrogen dioxide.

As with carbon monoxide, motor vehicles are the major source of the oxides of nitrogen in most parts of the country, although power stations and other large combustion units may be significant localised sources as well.

The main health impacts of the oxides of nitrogen come from nitrogen dioxide which is a respiratory irritant. At high levels it can corrode materials such as metals, and damage plants. Nitric oxide is believed to be quite harmless at the levels normally encountered in urban air.

The oxides of nitrogen are an important air pollutant because of their role in the formation of photochemical smog and of nitrates which take the form of fine particles and impair visibility. In the Northern Hemisphere, they also contribute to the problem of acid rain by converting to nitric acid in the atmosphere.

Most monitoring of the oxides of nitrogen (NOx) has been carried out in Auckland and Christchurch, and to a lesser extent, Dunedin. Two monitoring methods have been used: wet-chemical (bubbler) systems, which generally yield a 24-hour average result, and instrumental monitoring, which gives continuous data. Both systems produce results for both nitric oxide and nitrogen dioxide, although our primary interest here is with nitrogen dioxide.

Oxides of nitrogen were monitored using wet-chemical systems at 5 sites in Auckland over the mid- to late 1970s, at 10 sites in Christchurch for variable periods through most of the 1970s and 1980s, and at 2 sites in Dunedin in the late 1970s. Most of the sites were located either within or near the central business districts, or in suburban areas with significant motor vehicle activity.

The results can be summarised as follows, with all NOx data quoted in terms of nitrogen dioxide (NO2) equivalents:

  • In Auckland, maximum daily values for NOxwere typically in the range of 100 to 300 micrograms per cubic metre (100300 µg/m3), with annual means of 40 to 60 µg/m3.
  • In Christchurch, maximum daily values for NOx ranged from about 200 to 500 µg/m3, with occasional excursions above this limit. Annual means were in the range of 50 to 100 µg/m3in most locations.
  • In Dunedin, maximum daily values for NOxwere typically in the range of 150 to 300 µg/m3, with annual means of 50 to 60 µg/m3.
  • The ratio of nitric oxide to nitrogen dioxide in the above data was normally in the range of about 1:1 to 3:1. In other words, nitrogen dioxide normally accounted for between 25 percent and 50 percent of the total NOx. The main factor here is proximity to the sources, in that the ratio tends to drop with increasing distance as nitric oxide is converted to nitrogen dioxide in the atmosphere.

Monitoring for NOxwas carried out in a fairly sporadic fashion from the mid-to late 1980s, although there are now two permanent sites operating in Auckland and one in Christchurch.

Generally speaking, the instrumental data indicates slightly lower levels of NOxcompared to the wet-chemical information given above. This may be due to the improved sensitivity of the instrumental method or to a difference in location.

The above monitoring indicates there may be cause for concern at some sites in both Auckland and Christchurch for a small percentage of each year. However, because observations are based on levels of total NOx, more work is needed to determine whether nitrogen dioxide levels actually approach the guideline values.

Unlike some of the other pollutants mentioned in this report, levels of NOx do not appear to have changed significantly over the last 20 years. However, as with carbon monoxide, increasing vehicle numbers and use in recent years may change this.

Photochemical smog and tropospheric ozone

Photochemical smog is a dense brown haze which occurs in many large cities. It consists of a complex mixture of chemical gases, including ozone (O3). Although ozone is a vital radiation screen in the upper atmosphere, at ground level it is one of the main components of smog. Ozone is part of the mix of pollutants in photochemical smog, and is formed from reactions between other pollutants.

Photochemical smog results from the reaction of sunlight on the complex chemical mix of oxides of nitrogen (NOx)(which come from the oxidation of atmospheric nitrogen during and after the process of burning), and volatile organic compounds (VOCs) (emitted mainly from car exhausts, and to a lesser extent from industrial solvents).

Photochemical smog has recently been shown to increase death rates in big cities. A recent study in London found that ozone levels were associated with significant increases in respiratory and cardiovascular deaths (Anderson et al., 1996). This effect was quite independent of the effects of other pollutants (e.g. PM10). The mixture of chemicals present in smog is also extremely irritating to the eyes, nose, throat and lungs, and can cause breathing difficulties, particularly in susceptible people. It also causes deterioration of materials such as rubber, damages sensitive plants, and seriously reduces atmospheric visibility.

Because ozone is one of the main components of photochemical smog it is usually monitored as a simple indicator of smog conditions.

The only monitoring for ozone and/or photochemical oxidants in New Zealand, however, was carried out in Auckland almost 20 years ago. The city, then, was considered the only place where there were enough of the major emission sources (especially motor vehicles) and sufficient sunlight for smog to be a potential problem.

Photochemical smog received particular attention in the late 1970s and early 1980s when proposals were made to build a number of large power stations in and around the Auckland area. Most of the monitoring was carried out in joint studies by the Department of Health, the New Zealand Meteorological Service, and the New Zealand Electricity Division of the Ministry of Energy.

The results can be summarised as follows:

  • There is a natural background level of ozone which is normally in the range of 0 to 30 ppb.
  • Periods with higher levels of oxidants were observed in Auckland on 5 to 10 days of any year. These 'incidents' were generally in the range of 40 to 80 ppb (80 to 160 µg/m3), and usually occurred over periods of 2 to 3 hours in early to mid-afternoon during the summer months.

These results indicate that 20 years ago our cities had comparatively fewer smog 'incidents' relative to cities in other parts of the world. Although there has not been any recent monitoring for smog, the main contributors, particularly motor vehicles, have definitely increased, especially in Auckland. However, reliable evidence can only be obtained through renewed monitoring.


Lead is one of the group of so-called 'heavy metals' which also includes elements such as mercury, cadmium, and zinc. It is the most prevalent of these as far as air pollution is concerned. Leaded petrol was a widespread source of lead in the air in New Zealand, with the lead being emitted as fine particulate matter from motor vehicle exhausts. A few industrial sources, such as lead smelters and scrap metal recovery operations, may also contribute to local lead concentrations in the air. Although the main cause of lead-related illness in New Zealand is dust and lead-based paintwork in older buildings (see Box 6.2), lead in petrol has caused most of the widespread accumulation of lead in the environment.

Particulate lead has been monitored at more than 25 locations throughout the country, with sites in Auckland, Hamilton, Palmerston North, Hutt Valley, Wellington, Christchurch, and Dunedin. Two of these sites (Mount Albert and Penrose in Auckland) have been monitored since 1961, while many of the others have run for periods of 5 to 10 years.

Figure 6.11: Atmopsheric lead levels (annual averages) from five monitoring sites (two in Auckland, and one in Hamilton, and Christchurch) 1984-93

Annual averages for the atmospheric lead levels have decreased in all sites, most markedly since 1986, and levels in 1993 are between a third and a half of the levels in 1984. The decrease is directly linked to the gradual reduction of lead in petrol.

Source: Institute of Environmental Science and Research Ltd.

The monitoring sites cover a variety of locations, including suburban, inner-city, and alongside major motorways, with most of them directed at observing the effects of motor vehicle emissions.

Prior to 1986, when lead levels in petrol were reduced, concentrations in most urban areas were typically in the range of 0.1 to 1.0 µg/m3. In areas of high traffic densities or significant congestion the levels were generally higher than this, with values at times in excess of 2 to 3 µg/m3.

Since 1986, the lowering of the lead content in petrol has resulted in a marked reduction in atmospheric lead levels (see Figure 6.11 and 6.12).

Figure 6.12: Airbourne lead concentration (3-monthly moving averages) at two Auckland monitoring sites (Queen Street and Mt Eden), 1983-96

The 3-monthly moving average of airbourne lead levels has decreased in both Mount Eden and Queens Street with levels in 1996 having a concentration of around 0.1 micrograms per cubic metre. In 1983 the concentration in Queens Street was around 1.5 micrograms, peaking at nearly 2.5 micrograms in 1984. In 1988 the concentrations at Queen Street dropped below 1 microgram, the concentration recommended by the Air Quality Guidelines. The Mount Eden site has been below the recommended level throughout the time period.

Source: Institute of Environmental Science and Research Ltd.

Although the annual average statistics in Figure 6.11 clearly show the downward trend in atmospheric lead levels over recent years, they do not reflect marked seasonal variations recorded at the monitoring sites and shown in Figure 6.12. The data are presented for two Auckland sites as 3-monthly running averages for comparison against an air quality guideline of 1.0 µg/m3(3-monthly average).

The seasonal changes reflect variations in weather conditions. The highest lead levels are recorded in midwinter, probably because of the increased frequency of early morning temperature inversions in the winter months when air is trapped at ground level, below cloud or a warmer layer of air.