handbook of atmospheric science || urban-scale air pollution

15.1 INTRODUCTION

Urban regions are by their nature concentrationsof humans, materials, and activities. They there-fore exhibit both the highest levels of pollution andthe largest targets of direct impacts. Air pollutionis, however, enacted on all geographic and tempo-ral scales, ranging from strictly “here and now”problems related to various aspects of humanhealth and well-being, over regional phenomenalike acidification and forest die-back with a timehorizon of decades, to global phenomena, which inthe course of centuries have impacts for humansand nature over the entire globe.

Urban-scale air pollution is a vast subject, withdifferent socioeconomic aspects in different partsof the world —sometimes even within a specific re-gion. And it cannot be treated in isolation, butmust be seen in connection with air pollution onother scales —both when large-scale phenomenainfluence the urban air quality and when urbansources, via long-range transport, give rise to large-scale pollution impacts.

15.1.1 Urban air pollution as a chain of events

In a systematic treatment pollution can be con-ceived as a chain of decisions and processes. Anexample for urban-scale air pollution is shown inFig. 15.1.

The chain starts with a series of (often un-planned) decisions about activities related to spaceheating, traffic, industry, etc. These activities lead

to emission of pollutants, which are transported,dispersed, and sometimes transformed in the at-mosphere. Depending upon the conditions the re-sult is levels (concentrations) of pollutants thathave generally unwanted, impacts. The impactsare evaluated and result in attempts at mitigation,which may be applied at every link of the chain inthe form of reduction of emissions, facilitation ofdispersion, or counteraction of impacts.

On the urban scale there is often, but not al-ways, a reasonable relation between local emis-sions and resulting local pollution levels. Incontrast to regional and global pollution, efforts tomitigate urban-scale air pollution may thereforegenerally have direct and observable effects.

15.1.2 Air pollution through the ages

Initially air pollution was an indoor phenomenon,caused by open fires without controlled venting.As an example it has been shown (Skov et al. 2000)that the indoor exposure to nitrogen oxides and organic compounds in a reconstructed Iron Agehouse must have been comparable to, or even higher than, that in a modern city. Similar observa-tions have been made in present-day dwellings inthe developing world (Smith 1988, 1993). Ambienturban air pollution, however, is as old as cities, andliterature as well as historical records testify thatthe problems were extensive. Attempts at regula-tion have been documented for at least 2000 years(e.g. Mamane 1987). Some problems may evenhave been underestimated, since generally people

15 Urban-Scale Air Pollution

JES FENGER

Handbook of Atmospheric Science: Principles and ApplicationsEdited by C.N. Hewitt, Andrea V. Jackson

Copyright © 2003 by Blackwell Publishing Ltd

400 jes fenger

were less critical about their living conditions, andthey had no means of evaluating long-term im-pacts of, for example, carcinogens. Further, manyof the records concern aesthetic impacts in theform of smell and soiling, which are not deleteri-ous to health in themselves.

Finally, it should be recognized that up to theSecond World War many people had an ambivalentattitude toward pollution, which to some extentwas perceived as a symbol of wealth and growth.Thus advertisements showed pictures of fumingchimney stacks (Fig. 15.2) and automobiles withvisible exhaust —images hardly anyone would cul-tivate today!

Semiquantitative evaluations of early urban airpollution have been attempted in various ways,e.g. via records of material damage and impacts onhuman health and vegetation. Also, simplified dis-persion modeling is possible when the consump-tion of fuels and raw materials within a confinedarea is reasonably well known (Brimblecombe1977, 1987).

Some direct measurements of air pollutantswere carried out by scientists and amateur enthu-siasts in the nineteenth century, but systematicand official investigations with continuous timeseries are of fairly recent date.

15.1.3 Global growth and urbanization

Human societies as such were first formed withthe advent of agriculture about 10,000 years ago.The global population was then hardly more thanfive million. Until the start of industrializationand improved hygiene in the nineteenth centurythe population slowly grew to one billion, but thenthe rate increased dramatically. At the beginningof the twentieth century the number had doubledto two billion. Just after the Second World War the world population was about 2.5 billion, and inthe past 50 years it has more than doubled and hasnow passed six billion. Some projections suggestthat the number will level off at about 10 billion inthe second half of this century, with the sharpestrise in the (now) developing regions, especiallyAfrica.

Since 1950 global urbanization, defined as the

Activity(traffic)

Emission(NO)

Dispersiontransformation

(NO + O3 ÆNO2 + O2)

Pollution levels(μgm–3)

Exposure(hours)

Dose(level � time)

Impact(irritation)

O3

NO2

Fig. 15.1 Pollution is a chain of decisions and processes.This also applies to urban air pollution, and is hereshown for traffic emission of NO in a North EuropeanCity. In practice many such chains interact.

Urban-Scale Air Pollution 401

fraction of people living in settlements above 2000inhabitants, has risen from below 30 to 45%. In themore developed countries it is now around 75%and in the less developed 37% (Population Refer-ence Bureau 1999). In 1950 there were only eightcities with inhabitant numbers above five million,in 1990 there were about 35 (WHO/UNEP 1992),and now there are probably more than 40, abouthalf of them situated in East Asia. About 20 urbanregions each have populations above 10 millionpeople. By 2015 the number may have risen to 26,and for the foreseeable future the rate of megacitygrowth in the developing world will far outpacethat in the developed countries (Lynn 1999).

Today cities cover 2% of the Earth’s surface, andaccount for roughly 78% of the carbon emissionsfrom human activities, 76% of the industrial wooduse, and 60% of the water tapped for use by people(O’Meara 1999). These global trends will continuein the next decades (WHO/UNEP 1992). Particu-

Fig. 15.2 Advertisement for a British exhibition inCopenhagen 1932. Obviously before flue gas cleaningbecame fashionable.

7

6

5

4

3

2

1

0

1950 1960 1970 1980 1990 2000

World, rural

Total world population

Developing, urban

Industrial, urban

Billion

Fig. 15.3 The increase since 1950 of the total worldpopulation and the urban population in developed anddeveloping countries. (Based on material compiled inBrown et al. 1998.)

402 jes fenger

larly in the developing countries, there is signifi-cant migration of people from the countryside tothe towns —both because of the mechanization of farming and because of opportunities in newindustries and public services. This may lead to afurther growth in urbanization (UNEP 1997). InChina alone more than 100 million people havebeen reported to move around in search for work,and in Beijing the authorities appear to activelyprevent their permanent settlement. In Asia, LatinAmerica, and Africa this urbanization has been ac-companied by the proliferation of slums and squat-ter settlements (UN 1997a).

Regions with high birth rates and extensive im-migration are faced with environmental problemsdue to unplanned urban growth and emergingmegacities (Lynn 1999). In some cases the situa-tion is aggravated by polluting industries, whichhave been transferred from industrialized coun-tries with stricter environmental legislation andhigher wages. On the other hand there has been a tendency in the industrialized world that rela-tively fewer people live in the inner areas of the cities (OECD 1995b). The consequences are ex-panding road systems and increasing commutingtraffic, which may counteract improvements inautomobile technology.

15.1.4 Global energy consumption, production, and emissions

World production of automobiles has increased bya factor of five since 1950 (Fig. 15.4) and the totalfleet is now above 500 million. In addition to thisthere has been a substantial production of motor-bikes, nearly half in China, with an estimated fleetof 80 million (OECD 1995a).

In the same period industrial production has in-creased by a factor of 10 and global energy con-sumption by nearly a factor of five (UN 1997b andearlier reports). Since the major part of the energyhas been produced by fossil fuels, and to a minorextent by biofuels, initially without flue gas clean-ing, the global emissions of air pollutants haveincreased correspondingly. Since 1950 the globalemission of sulfur oxides has more than doubled,and the emission of nitrogen oxides increased by a

factor of four. A substantial part of these emissionsoccur in urban areas.

15.2 POLLUTANTS AND SOURCES

To date about 3000 different anthropogenic air pol-lutants have been identified, most of them organiccompounds (including organometals). Combus-tion sources, especially motor vehicles, emit sev-eral hundred different compounds, but only forabout 200 of them have the impacts been investi-gated. Ambient concentrations are known for amuch smaller number and time series for evenfewer.

The pollutants can be divided into two groups(Wiederkehr & Yoon 1998): the traditional majorair pollutants (MAP, comprising sulfur dioxide,particles, nitrogen dioxide, carbon monoxide, lead,and ozone); and the hazardous air pollutants (HAP,comprising chemical, physical, and biologicalagents of different types). The HAPs are generallypresent in the atmosphere in much smaller con-centrations than the MAPs, and they often appearmore localized (typically in urban areas or near in-dustries), but they are —due to their high specificactivities —nevertheless toxic or hazardous. Bothin basic investigations and in abatement strategiesHAPs are difficult to manage, not only because of

1950 1960 1970 1980 1990 2000

40

30

20

10

0

Million

Motorbikes

Automobiles

Fig. 15.4 The world production of automobiles (uppercurve) and motorbikes (lower curve). (Based on materialcompiled in Brown et al. 1998.)


their low concentrations but also because they arein many cases not identified.

Compounds that increase the greenhouse effect(carbon dioxide, methane, nitrous oxide) or depletethe ozone layer (e.g. CFCs) are not urban pollutantsper se, but control of their emissions is related toemission of other pollutants, and their impactshave a bearing on urban quality of life.

15.2.1 Characteristics of pollutants

Sulfur dioxide (SO2) is the classical air pollutant as-sociated with sulfur in fossil fuels. Emission can besuccessfully reduced using fuels with low sulfurcontent, e.g. natural gas or oil instead of coal. Inlarger plants in industrialized countries desulfur-ization of the flue gas is an established technique.

Nitrogen oxides (NOx) are formed by oxidationof atmospheric nitrogen or nitrogen compounds inthe fuel during combustion. The main part, espe-cially from automobiles, is emitted in the form ofthe nontoxic nitric oxide (NO), which is subse-quently oxidized in the atmosphere to the second-ary and toxic pollutant NO2. Emissions can bereduced by optimization of the combustionprocess (low NOx burners in power plants and leanburn motors in motor vehicles) or by means of cat-alytic converters in the exhaust.

Carbon monoxide (CO) is the result of incom-plete combustion. Emissions can be reduced by in-creasing the air/fuel ratio, but with the risk ofincreasing the formation of nitrogen oxides. In automobiles the most effective reductions are carried out with catalytic converters.

Particulate matter (PM) is not a well defined en-tity. Originally it was determined as soot or “blacksmoke,” for which there is a European Union (EU)air quality limit value (Edwards 1998). Later theconcept of total suspended particulate matter(TSP) was introduced, but since 1990 size fraction-ating has been attempted by measurements ofPM10 (particles with aerodynamic diameter lessthan 10mm). The major part of PM10 may have anatural origin (e.g. sea spray or desert and soil dust),and it is therefore important to also measure PM2.5or even, when the appropriate technology has beendeveloped, PM1.

Usually particles are grouped in three modes:ultrafine, fine, and coarse. The ultrafine particlesare formed by chemical reactions or are condensedfrom hot vapor, e.g. from diesel exhaust, and theycoagulate into fine particles (Whitby & Sverdrup1980). Defined as having an aerodynamic diameterless than 0.1 and 2.5mm respectively (UNEP/WHO1994), the ultrafine and fine particles are thus pre-dominantly of anthropogenic origin. Coarse parti-cles, on the other hand, are often of natural origin(dust, seaspray, pollen, or even insects). In determi-nations of TSP the coarse particles dominate withtheir high mass, almost irrespective of their rela-tive number.

In urban atmospheres the actual size spectrashow quantitative differences with, for example,more pronounced mass peaks for fine particles insuburban sites and for coarse particles near seacoasts (Fig. 15.5).

The application of different measuring tech-niques complicates the evaluation of changes inpollution levels. To some extent it is possible toestablish relationships between the concentra-tions of fine and coarse particles relevant to epi-demiological studies (Wilson & Suh 1997), butonly under well defined conditions. Measure-ments from Erfurt in the former East Germany

Total mases of airborne particles

0.3 0.5 1.0 2 5 10 20 50 100

Mesoscale/suburban

Sea coast site

Urban industrial avarage

Urban (traffic) industrial

Aerodynamic diameter (mm)

Fig. 15.5 Schematic size distribution of particulatematter in various atmospheres. (From UNEP/WHO1994). The corresponding deposition in the humanrespiratory system is shown in Fig. 15.13.

404 jes fenger

show that the level of PM2.5 has been reduced sub-stantially since the reunification of Germany andthe subsequent introduction of updated technolo-gy. Nevertheless, the amount of even smaller par-ticles has increased and so has the total number ofparticles, indicating a change in major sources(Tuch et al. 1997).

A further complication is that the chemicalcomposition of particles is not well known, andthat health impacts may be due to other pollutantsadsorbed on them —typically heavy metals or lessvolatile organic compounds.

Emissions of particles of anthropogenic origincan be reduced by use of cleaner fuels, better com-bustion techniques, and a series of filtration or im-paction technologies.

Volatile organic compounds (VOCs) as air pol-lutants are the result of incomplete combustion offuels, are formed during combustion, or are due toevaporation. Some industrial processes and the useof solvents result in the emission of VOCs. Inurban air the most important compounds are ben-zene and the series of polyaromatic hydrocarbons(PAH), but 1,3-butadiene, ethene, propene, and aseries of aldehydes have also received attention(Larsen & Larsen 1998).

Biogenic VOCs, emitted from vegetation, donot pose a health risk in themselves, and thesources in cities are modest, but they must betaken into account in relation to regional photo-chemical air pollution, which in turn may influ-ence urban air quality.

Lead as an additive to gasoline has by and largebeen phased out in the major part of the industrial-ized world, but it is still used in many developingcountries and economies in transition, whereemissions from industrial activities are also signif-icant. Other heavy metals of interest as air pollu-tants include cadmium, nickel, and mercury, allwith industrial sources.

Ozone is a secondary pollutant formed in pho-tochemical reactions between VOCs and nitrogenoxides. Apart from small sources of indoor pollu-tion like photocopiers and printers, it has no directsources, and it can therefore only be controlled via the primary pollutants. Other constituents ofphotochemical air pollution, e.g. aldehydes and

PAN (peroxy-acetyl-nitrate), are not regulatedseparately.

15.2.2 The main source groups

Since air pollution is mostly related to combustionor evaporation, the different sources emit by andlarge the same compounds but in varying propor-tions, under different conditions, and with differ-ent time patterns (Stanners & Bourdeau 1995).There is also geographic variation, thus, for exam-ple, in eastern Europe SO2 from space heating playsa relatively more important role compared to west-ern and southern Europe, and in southern Europethe contribution of SO2 from traffic is relativelyhigh due to the use of diesel oil with a high sulfurcontent.

Space heating from small individual units wasoriginally the main source of SO2, CO, soot, andorganic compounds in cities, but now —with theintroduction of cleaner fuels, better combustiontechnology, and district heating —has diminishingemissions in industrialized countries.

Power generation with fossil fuels, in somecases combined with district heating, is, despitecleaner fuels and flue gas cleaning, still an impor-tant source of SO2 (more than half) and NOx (aboutone-fifth) in Europe. In general, however, the im-pacts on urban air quality are relatively modest,since many plants are located in rural areas andequipped with high stacks.

Incineration of waste in larger plants has somesimilarities to power production, and is in manycases used for district heating. Due to the verymixed fuels, the emission of heavy metals andtoxic organic compounds (e.g. dioxin) must beconsidered.

Industry is indirectly responsible for all theemissions related to energy use. Depending uponthe production, emission of organic compoundsand heavy metals may be significant. If the emis-sions arise from diffuse sources they may be diffi-cult to control and have relatively large localimpacts.

Use of solvents in households, crafts (painting),and minor industries gives rise to the evaporationof organic compounds. A more responsible use, in-


cluding a switch to water-based coatings, can byand large solve the problem.

Traffic, and especially individual motorizedtraffic in the form of automobiles and motorbikes,is the dominant source of air pollution in mosturban areas (Schwela & Zali 1999), not only interms of local emissions, but also in terms of re-sulting pollution levels, since the emissions takeplace at low height and often in street canyons. InEurope all mobile sources account for nearly 70%of the emissions of CO, more than 60% of the NOx,and more than 30% of the VOCs. Polyaromatic hy-drocarbons (PAH) are mainly emitted from traffic(Nielsen 1996; Nielsen et al. 1999) and some haveonly recently been noticed (Enya et al. 1997).

The almost complete removal of lead as an addi-tive to gasoline (Nriagu 1990) has not been com-pletely without risk of side-effects. Changes in thecomposition of the gasoline to increase the octanenumber may increase the emission of aromatichydrocarbons, including benzene. Benzene con-centrations have increased in many urban at-mospheres with the introduction of catalyticconverters (Richter & Williams 1998). An alterna-tive additive, MTBE (methyl-tert-butyl ether), notonly increases the octane number, but also im-proves the combustion and thus reduces theemissions of carbon monoxide and hydrocarbons.It is, however, an air pollutant, causing bothimmediate eye and respiratory irritation and long-term risk of cancer. More important may be thecontamination of soil and groundwater, especiallyaround gasoline filling stations (transmedia pollu-tion). In Denmark MTBE is only used for 98 octanegasoline.

Particles in the urban atmosphere, and espe-cially small particles from diesel exhaust, are re-ceiving increasing attention (Marico et al. 1999;Maynard & Howard 1999; Schwela & Zali 1999).

15.2.3 Emission inventories

Emissions are measured or calculated individuallyonly for large point sources, such as power plants.In most cases emission inventories are carried outon the basis of emission factors. These factors mayexpress the emitted amount of a given compound

for a given activity or use of fuel. A typical emis-sion factor might thus be: “0.013kg CO GJ-1, whennatural gas is used in district heating plants” (fordetails see Chapter 17).

National inventories of emissions as they are carried out, for example, in the EuropeanCORINAIR database (CORINAIR 1996) are usedin international negotiations and may for lack ofbetter information suggest general trends in airpollution levels. For studies of transboundary airpollution a spatial resolution of 50km is normallysufficient (EMEP/MSC-W 1998, 1999).

For urban areas, however, more detailed inves-tigations are necessary, with time and space reso-lutions relevant to the applied scale. Proxy data for larger areas can be generated on the basis ofinformation on, say, traffic patterns (Friedrich &Schwarz 1998), but detailed investigations of indi-vidual streets (Berkowicz 1998) must be based onactual traffic counts. Of special importance is thetime-dependent traffic density, with more or lesspronounced rush hours (Fig. 15.6).

15.3 FROM EMISSION TOPOLLUTION LEVELS

Emission of air pollutants from urban sources de-termine to a large extent the urban air quality.However, advection, dispersion, and to some ex-tent simultaneous transformation are also impor-tant. Therefore, the impacts of different sourcesand their relative importance cannot be evaluatedon the basis of emission inventories alone.

In the design of cost-effective abatement strate-gies (e.g. Krupnick & Portney 1991) it must be real-ized that the relations between emissions andresulting concentrations (so-called “immissions”)are by no means simple. Measurements are stillthe foundation of understanding, but applicationof mathematical modeling, and also of physicalmodeling in wind tunnels, is of increasing impor-tance in urban air pollution management. As aconsequence, numerous techniques have beendeveloped for different spatial scales, ranging fromentire regions down to individual streets. Somemodels only describe dispersion or have simple

406 jes fenger

reaction schemes; more sophisticated modelscomprise a large number of interacting reactions.Such models can be further developed to form full decision support systems (e.g. Dennis et al.1996).

15.3.1 The urban climate

An urban area differs from the surrounding ruralregion in important ways, which influence its cli-mate, the possibility of dispersion of air pollutants,and their impacts. A city is generally darker thanits surroundings and thus has a higher energy ab-sorption (lower albedo). This effect is amplified bythe rougher surface created by buildings, whichtrap solar radiation. Finally, various energy-consuming activities add to the heating, whichmay result in temperatures a few degrees above thesurrounding rural areas.

At low wind speeds this so-called “urban heatisland” can give rise to a circulation where the pol-lution is trapped. This situation can be aggravatedby the topography, e.g. if the city is situated in avalley, which further restricts horizontal mixing.At higher wind speeds the urban area acts as thesource of a plume, with elevated pollution levelsdownstream (Fig. 15.7).

The relative humidity in the city center is gen-erally lower —partly because of the elevated tem-perature, partly because of enhanced runoff ofprecipitation in sewers and drains. On the otherhand, the general instability generated by the heat

island and the turbulence due to the rough surfacemay increase rainfall (Goldreich 1995).

15.3.2 Urban-scale dispersion andtransformation

Dispersion in the urban area

The importance of dispersion was recognized withthe invention of chimneys, and meteorologicalconditions played a crucial role in a series of pollu-tion disasters (Brimblecombe 1987). Dispersionmechanisms have received special interest withthe increasing traffic in built-up areas, wherestreet canyons exhibit special flow patterns (Fig.15.8). Consistent with this, both measurementsand dispersion calculations (e.g. Berkowicz 1998)have shown that the wind direction is importantfor long-term pollution levels; therefore, in areaswith a dominant wind direction – in northernEurope from the west – the orientation of the indi-vidual street may result in significantly differentlevels at the two sides. In streets parallel to thewind direction traffic emissions may give concen-trations which built up downstream.

Wind speed is also of importance. Figure 15.9demonstrates the significance of wind speed forthe resulting air pollution in a street canyon,where a fairly strong wind (8ms-1) is seen to nearlyhalve the concentration of NO2 at rush hours.

The overall significance of the climatologicalconditions is clearly demonstrated in a com-

Share

of

tota

l daily

tra

ffic

load (

per

hour)

Athens, Madrid

Cologne, London. The Hague

Lyon

Milan

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

00 2 4 6 8 10 12 14 16 18 20 22 24

Fig. 15.6 Hourly profiles of trafficdensity for a series of Europeancities in 1990. (From EuropeanCommission 1996b.)


parison between air quality in Copenhagen andMilan. Since the frequency of low wind speeds isconsiderable higher in Milan than in Copenhagen,Milan has much higher pollution levels for compa-rable emissions (Vignati et al. 1996). In recent stud-ies by Cocheo et al. (2000) the urban levels ofbenzene measured in six European cities werefound to decrease drastically with increased windspeed.

Chemical reactions

During dispersion the pollutants interact chemi-cally (e.g. Finlayson-Pitts & Pitts 1997) and for theurban atmosphere reactions between nitrogen ox-ides, organic compounds, and ozone are the mostimportant. Photochemical smog with formationof ozone and other oxidants was first recognized inLos Angeles in the mid-1940s as an urban phenom-

Rural Suburbs SuburbsCity center

Urban boundary layer

Urban canopy sub-layer

Rural

V >3ms–1

Urban "dome" model

Urban dome

Suburbs City center Suburbs RuralRural

(b)

V>3ms–1

Urban "plume" model(a)

Fig. 15.7 Urban heat island in thecase of (a) moderate to strong windand (b) weak wind. (From Mestayer& Anquetin 1995; with kindpermission from Kluwer AcademicPublishers.)

Roof level wind

Background pollution

Recirculating air

Direct plume Windwardside

Leewardside

Fig. 15.8 Vortex in a street canyonwith a roof level wind at a rightangle to the direction of the street.Even at moderate wind speeds (3–4ms-1) short-term pollutionconcentrations may be 5–6 timeshigher at the leeward site. For winddirections parallel to the street, the“vortex effect” disappears, but maybe replaced with a “tunnel effect,”where concentrations build up inthe wind direction.

408 jes fenger

enon related to automobile exhaust in a subtropi-cal, topographically confined region.

Photochemical smog is now observed in manyparts of the world, but with distinctly different pat-terns. In the south of Europe, cities like Athens andRome may experience a “summer smog” of the LosAngeles type, but in many cases it is a large-scalephenomenon (Guicherit & van Dop 1977). In citiesin the northern part of Europe, one of the predomi-nant reactions is a reduction of ozone by nitricoxide in automobile exhaust (Fig. 15.1) to formoxygen and nitrogen dioxide:

As a result ozone levels are generally lower atground level in the streets than at roof level or inthe surrounding countryside. Urban ozone levelsare higher during weekends with low traffic andmay be practically nil during some pollutionepisodes (Fig. 15.10). Note also in Fig. 15.10 thatthe concentration of NO follows the traffic in-tensity at rush hours and weekends much moreclosely than NO2, the concentration of which islargely determined by the available O3, suppliedfrom outside the city. A disappointing conse-quence of this reaction pattern is that a given re-duction in the emission of nitrogen oxides (NOx),e.g. by the introduction of catalytic converters onautomobiles, will generally result in less than thecorresponding reduction in the concentrations ofNO2 (Palmgren et al. 1996).

These mechanisms often lead to the formationof elevated ozone concentrations downstreamfrom the city (city or urban plume). In a similarphenomenon on a larger scale, elevated ozone con-centration in central Europe due to extended high-pressure events can, via long-range transport, bedetected in northern Europe. Here concentrationsnormally build up over several days —often in par-allel with rising temperatures, high stability, andsolar radiation.

Since ozone is a secondary pollutant it can onlybe regulated via the primary pollutants. Long-range chemical/transport models can demonstratethe effects of changes in the emissions and ozoneconcentrations, and they clearly indicate that a

NO O NO O3 2 2+ Æ +

2500

2000

1500

1000

500

0

10

8

6

4

2

60

40

20

0

September 12 September 13

NO

2 (

p.p

.b.)

Win

d s

pee

d (

ms–

1)

Veh

icle

s per

hour

6 12 18 24 6 12 18 24

Fig. 15.9 Relations between traffic intensity, windvelocity, and NO2 concentrations on two consecutiveweekdays in 1997, measured at Jagtvej, a busy streetsurrounded with buildings in Copenhagen. Theconcentrations are measured both at the roadside (uppercurve) and at a nearby rooftop (lower curve). (Based onmaterial from Kemp et al. 1998.)


concerted international effort is necessary. In com-puter experiments it has thus been shown that inthe hypothetical situation where all Danish emis-sions were reduced to zero, the average ozone

levels in Denmark would go up slightly (Zlatev et al. 1996).

Formation of particles

As stated in Section 15.2.1 particles in urban aircan have many forms. Primary particles can befairly large soot particles that cause soiling(Section 15.4.2), but are of minor importance forhuman health (Section 15.4.1). Less volatile organ-ic compounds from, in particular, diesel exhaustmay condense and form smaller particles or mayadd to small soot particles. The actual formation of secondary particles is predominantly a result ofoxidation of sulfur dioxide to sulfate and nitrogenoxides to nitrate. The most important oxidizingagent is hydroxyl, OH-radicals being formed inphotochemical reactions. Sulfuric acid vapor nu-cleates on its own or with water molecules to forma fine aerosol in the nanometer range. Nitric acidmay, for example, react with ammonia and formammonia nitrate particles. Although these sec-ondary particles may be formed in the urban at-mosphere they are spatially more uniformlydistributed, with smaller urban to rural gradientsthan primary particles in traffic emissions.

15.3.3 Relations between air pollution on different scales

In more open spaces (parks, squares, residentialareas) the pollution levels take the form of anurban background, with increasing impacts frommore distant sources. Recent applications ofmesoscale computer models have also demon-strated that the regional component is important,especially in areas with a complex landscape such as coastal regions; thus studies in theMediterranean region and southern Europe haveindicated that in certain periods the urban areasmay be significantly affected by sources locatedhundreds of kilometers away (Kallos 1998).

Figure 15.11 shows as an example and symboli-cally the average impacts of different sources onthe urban levels of nitrogen oxides.

Growing global consumption of fossil fuelsleads to energy-related emissions of carbon dioxide

80

60

20

0

40

160

120

80

40

0

800

600

400

200

0M T W T F Sa Su

NO

(mg

m–

3)

NO

2 (

mgm

–3)

O3 (

mgm

–3)

Fig. 15.10 Average weekly variations for ozone andnitrogen oxides at H.C. Andersens Boulevard inCopenhagen. For comparison a typical variation at arural site is shown. (Based on HLU 1994.)

410 jes fenger

(e.g. Ellis & Tréanton 1998) and may eventually,via the enhanced greenhouse effect, result in cli-mate changes with impacts on all human activi-ties and natural ecosystems. One of the results ofthe UN conference on environment and develop-ment in Rio de Janeiro in 1992 was an action planfor the attainment of a sustainable global develop-ment in the twenty-first century —the so-calledAgenda 21. As a consequence many cities and ad-ministrative units in the industrialized world haveembarked on local programs, and more than 290European cities have signed the Aalborg Charter ofEuropean Cities and Towns towards Sustainabili-ty. Noteworthy in this connection is The Interna-tional Council for Local Environmental Initiatives(ICLEI 2000), whose purpose is to achieve andmonitor improvements in global environmentalconditions through cumulative local actions. Al-though political attention emphasizes climateprotection and thus reduction of emission of theultimate product of combustion, CO2, attempts tosave energy may also improve urban air quality(e.g. Pichl 1998).

On the other hand, climate changes may alterthe dispersion of urban air pollutants via a change(possibly an increase) in the occurrence of anticy-clonic conditions, during which the dispersion islimited. This possibility is discussed by the Inter-governmental Panel of Climate Change (IPCC),but no definite conclusions have been reached sofar. Clearly, the links between climate change,urban air quality, and human exposure to air pollu-tants are extremely complex and cannot as yet bequantified.

15.3.4 Outdoor and indoor air quality

Indoor air quality is largely determined by indoor sources —not only in occupational environ-ments, but also in offices, libraries, museums(Brimblecombe 1990), and private homes withemissions from open fires, evaporation from syn-thetic building materials and solvents, use of de-tergents, smoking, etc. However, in urban areasambient air pollution often has a significant influ-ence, which must be taken into account in theevaluation of various impacts.

Generally the contributions from outdoorsources are reduced, and especially in buildingswith recirculation and fresh air intake throughfilters. In a study in the National Museum ofDenmark (Schmidt et al. 1999) it was found thatelements of mainly antropogenic origin (sulfur,vanadium, and lead, and elements of mainly soilorigin silicon, cadmium, and titanium) have thesame time pattern in an unventilated room as inthe outdoors, but with indoor concentrations re-duced to about one-fifth. In a ventilated room thelevels were reduced to about 1%. Nitrogen dioxidewas reduced to one-half in the unventilated roomand to one-third in the ventilated.

A crucial issue is the impact of outdoor parti-cles on indoor particle concentrations. Theoreticalcalculations (Wallace 1996) indicate that withoutindoor sources and for a typical air exchange rate of0.75h-1 the indoor/outdoor ratio should be 0.65 forfine particles and 0.43 for coarse.

Fig. 15.11 A cut through an urbanarea with contributions to theconcentration of nitrogen oxides atground level. In cities in northernEurope the concentration of ozonevaries inversely with that ofnitrogen oxides.


15.3.5 Air quality indicators

The complex nature of air pollution, especiallywith respect to health impacts in cities, hasprompted attempts to define so-called indicators(Wiederkehr & Yoon 1998), which condense andsimplify the available monitoring data to makethem suitable for public reporting and decision-makers. The OECD (1998) has applied major pollu-tants measured in a specified way as indicators forthe total mix of pollutants.

In another type of synthesis the OECD (1998)has aggregated monitoring data from various re-gions (western Europe, USA, Japan) to demon-strate overall trends in pollution levels. Theresults of such an exercise should be treated withsome caution, and in particular comparisons be-tween different cities are dangerous (see Section15.6.1). Yearly averages, however, appear to repre-sent the general developments in specific citiesreasonably well (see Fig. 15.27).

The weighed means of concentrations of sever-al pollutants relative to guideline values have alsobeen used (Kassomenos et al. 1999). Again compar-isons between different cities are dangerous —ifnot impossible —due to differences in the mix ofpollutants, in the monitoring networks, and in theclimate-related impacts.

15.4 URBAN-SCALE IMPACTS

Urban air pollution has a series of impacts onhuman health and well-being, materials, vegeta-tion (including urban agriculture), and visibility.These impacts depend in the first instance on therelevant pollution levels, but also on other factors,such as climate, lifestyle, and the possibility of in-teraction between different components. For shortrecent reviews with references see Fenger et al.(1998, Chapters 18–21).

15.4.1 Human health and well-being

The impacts on human morbidity and mortality of air pollution have been unambiguously docu-mented during acute episodes in the past. Today

impacts on human health and well-being are moresubtle —at least in developed countries —but theyare still the main concern with urban air pollution.They also by and large determine abatementstrategies (Bascom et al. 1996; Holgate et al. 1999;WHO 2000).

In some cases the abatements have had impres-sive effects. Thus the phasing-out of lead in gaso-line in the industrialized countries (Nriagu 1990)has resulted not only in a drastic reduction inambient lead concentrations, but also in signifi-cant decreases in blood lead levels (typically by afactor of three) in the relevant populations, con-verging to a “natural level” of 3mgdl-1 (Thomaset al. 1999).

It should be emphasized, however, that thehealth of urban populations is determined bymany other factors related to urban living, whichblur the picture (Phillips 1993). The most impor-tant confounder is active and passive smoking,which has large regional and cultural variability,but in general appears to have health impacts wellabove those of general air pollution.

The human respiratory system

Air pollutants can have impacts on humansthrough various routes, e.g. when heavy metals aretransferred to food and water. Some compounds(ozone, nitrogen dioxide, aldehydes) act directly aseye irritants, but the dominant route is via the res-piratory system (Fig. 15.12).

For gaseous pollutants the rate of uptake de-pends upon solubility. SO2 is soluble and normallymore than 95% is deposited in the upper airways,whereas NO2 and O3 penetrate deeper into thelungs. The penetration of insoluble particles (soot,flyash, mineral dust) is highly dependent upontheir size (Fig. 15.13).

Even for a well defined exposure the uptake de-pends on respiration rate, which increases withphysical activity. The eventual effects dependupon the individual sensitivity, with the elderly,children, and asthmatics being especially sensi-tive to acute respiratory impacts.

412 jes fenger

Exposure

Until recently most studies of air pollution werecentered on determination of ambient levels,which have formed the basis of legislation andabatement strategies. Equally important, how-ever, is the extent to which people are actuallyexposed to the measured or calculated pollutionlevels. So far the pollution exposure of the popula-tion has in many cases been assessed on the basis ofcrude assumptions, e.g. that the levels observed at a single or a few monitoring stations are repre-sentative of the exposure of the population in a

larger urban area. Some results of such an evalua-tion (WHO 1995) are summarized in Section15.6.3.

Direct personal monitoring demonstrates thatthe levels an individual is exposed to vary drasti-cally during the day, and that the ambient, outdoorair quality does not adequately describe the actualexposure in sufficient detail (an example is shownin Fig. 15.14). A further complication is that peopleappear to stir up “personal clouds” of particle-laden dust from their surroundings and thereforemay experience exposure to fine particles about60% greater than classical monitoring wouldsuggest (Renner 2000).

A realistic evaluation of population exposuretherefore requires statistical information, wherethe time and activity pattern of the entire popula-tion must be taken into account. Thus Danishstudies (Andersen 1988) indicate that on averagethe population spent only one hour per day out-doors, one hour commuting and the remaining 22hours indoors. More recent studies (Jensen 1998)apply a Geographic Information System (GIS) tocombine air pollution data calculated with theOSPM model (Berkowicz 1998) with populationdata available from administrative databases.

Nose

Throat

Trachea

Primarybronchi

Secondary bronchi

Terminal bronchi

Lungs

Alveoli

Fig. 15.12 Simplified structure of the human respiratorysystem.

100

80

60

40

20

0

0.001 0.01 0.1 1 10 100

Ultrafine Fine Coarse

Dep

osi

tion (

%)

Upperairways

Total

AlveoliBronchi

Particle diameter (mm)

Fig. 15.13 Deposition of particles in the humanrespiratory system. Note that coarse particles arepredominantly caught in the upper airways, whereasfine and especially ultrafine particles penetrate deepinto the lungs.


In view of the large fraction of time spent in-doors, especially in cities with a cold climate, therelations between outdoor pollution and relatedindoor levels (see Section 15.3.4) must be takeninto account. That is not always simple; as de-scribed in Section 15.3.2, the general pollutionlevel for given emissions decreases with increasingwind speed. For benzene, however, this is not re-flected convincingly in personal exposure (Cocheoet al. 2000), and in some cases population expo-sures were found to exceed the average ambienturban concentration. It is speculated that the rea-son may be that people are generally outdoorswhen the ambient levels are high and indoorswhen they are low, and that the indoor environ-ment may store pollution with an outdoor origin (asort of “flywheel effect”).

Particles

Soot and TSP have always been considered a healthrisk, but in the past decade fine particles havereceived special attention —partly because of in-creasing traffic, partly because of the developmentof adequate monitoring techniques (Pope et al.1995; Maynard & Howard 1999; Section 15.2.1).

In an often cited investigation of the associationbetween air pollution and mortality in six UScities Dockery et al. (1993) reported a strong corre-lation between the concentrations of small parti-cles (PM2.5) and mortality rate ratios (Fig. 15.15).For other pollution parameters (TSP, SO2, sulfate,

O3, and acidity) some increase with pollutionlevels was observed.

In a larger study, Pope et al. (1995) linked ambi-ent air pollution data from 151 US metropolitanareas to the survival or death of 0.55 million per-sons. After correction for a series of confounders asignificant association between fine particulateexposure and survival emerged. Adjusted relativerisk ratios had a value of 1.17 between the highestand lowest polluted area.

In view of the surprisingly large effect the re-

5000

4000

3000

2000

1000

020:20 20:30 20:40 20:50 21:00 21:20 21:30 21:4020:10

Time of day

PC

(ng

m–

3)

aver

age:

49

2ng

m–

3

Fig. 15.14 Particle concentrationmeasured on a trip from AeropuerteInternacional to a hotel in the southof Mexico City. The measurementswere based on photoelectriccharging and emphasize polycyclicaromatic hydrocarbons (PAH) fromcombustion of organic material.(Reprinted from AtmosphericEnvironment (Zhiqiang et al. 2000).Copyright 2000, with permissionfrom Elsevier Science.)

1.4

1.3

1.2

1.1

1.0

0.9

0 5 10 15 20 25 30

Rate

rati

o

Fine particles (mgm–3)

Fig. 15.15 Estimated adjusted mortality-rate ratios andfine particle pollution levels (PM2.5) in six US cities. The rate for the least polluted city is defined as 1. (From Dockery et al. 1993.)

414 jes fenger

sults of these investigations should not be uncriti-cally transferred to other countries with differentage distribution, lifestyle, climate, and pollutionmix. Nevertheless, taking together the two stud-ies, the WHO estimates a relative risk of 1.1 perlong-term exposure to 10mg PM2.5 m-3. With aseries of further assumptions, Brunekreef (1997)calculates a corresponding loss in life expectancyof more than one year for Dutch men.

Numerous other epidemiological studies ofshort-term and long-term effects of air pollutionhave shown that fine particles at the present levelsare responsible for significant pulmonary impacts,especially for people already suffering from respi-ratory and cardiopulmonary diseases. Schwartz etal. (1996) have reported that a 10mgm-3 increase inPM2.5 two-day mean was associated with a 1.5%increase in total daily mortality. Somewhat largerincreases were found for deaths caused by chronicobstructive pulmonary disease (3.3%) and is-chemic heart disease (3.3%).

Notwithstanding this extensive documenta-tion, some important problems remain to besatisfactorily elucidated. First, the statistical pro-cedures are debated. The question of how to relatedaily mortality with longer-term mortality effects(McMichaelet al.1998) and especially how to filterout a possible “harvesting effect” (Zeger et al.1999) arises. Second, the assessment of human ex-posure is still being developed (Mage et al. 1999).Finally, it must be admitted that epidemiologicstudies provide little (or no) information on theunderlying impact mechanisms, which are stillpoorly understood (Brunekreef 1999).

Some recent European studies

In the APHEA (Air Pollution and Health, a Euro-pean Approach) the effects of several air pollutantsin a total of 15 European cities in 10 countries wereinvestigated. In western European cities it wasfound that an increase of 50mgm-2 in sulfur dioxideor black smoke was associated with a 3% increasein daily mortality. For PM10 it was 2%. The rises incardiovascular and respiratory mortality were 4and 5% respectively (Katsouynnia et al. 1997).

The daily hospital admissions for respiratory

diseases (Anderson et al. 1997; Spix et al. 1998)were found to increase, most significantly, forelevated levels of ozone, although the underlyingstatistics have later been criticized (Hasford &Fruhmann 1998).

In the Swedish SHAPE (Stockholm Study onHealth Effects of Air Pollution and Their Econom-ic Consequences) (Bellander et al. 1999) it has beencalculated how many hospital admissions can beavoided in the county of Stockholm if all levels ofPM10 and nitrogen dioxide are reduced to the levelsin the fringe areas (reduction of 5 and 16mgm-3 re-spectively). On this basis it is estimated that 700 ofthe yearly admissions of about 65,000 for respira-tory, cardiac, and circulatory diseases could beavoided.

It should be noted that Stockholm is a fairlyclean city (similar to the least polluted in the USstudies) and a complete removal of all particleemissions would only increase the average life-span by about 2 months. Nevertheless this is aboutdouble the effect of a hypothetical prevention of alldeadly traffic accidents.

In the so-called EXPOLIS (Air Pollution Expo-sure Distributions of Adult Urban Populations in Europe) study (Jantunen et al. 1998), personalexposure was determined by monitors and diaryrecords in six European cities (Helsinki, Bilthoven,Prague, Basel, Grenoble, and Milan).

Economic evaluation of health impacts

Cost–benefit analyses including valuation ofhuman health are controversial, with moral andethical complications. Nevertheless, this has beenattempted in many cases related to air pollution —often with the explicit aim of demonstrating thatcontrol is worthwhile considering the reduced lossof labor force, cost of health care, etc. In a recentstudy the European Commission (1998) performedan economic evaluation of the air quality targetsput forward in 1996 (see Section 15.5.1). The over-all conclusion was that in the case of PM10 the ben-efits exceed the costs by a factor of 100–200, thatfor SO2 and NO2 the benefit–cost ratios are in rangeof 1–10, but that in the case of lead the costs appearto exceed the benefits.


15.4.2 Material damage

The most conspicuous impact of urban air pollu-tion is the degradation of materials, which leads todirect economic loss, failure of equipment, and de-terioration of irreplaceable cultural artifacts. Thelocal urban climate (Section 15.3.1) is important,since temperature and relative humidity controlthe moisture layer on surfaces in the absence ofprecipitation. This effect partially counteracts thehigher pollution levels in the center of a city(Kucera & Fitz 1995). This can also explain thenoticeable impacts in fairly clean Nordic citiescompared to cities in the south.

Soiling

Particulate pollution has different effects: the soil-ing itself represents an aesthetic loss. Cleaning canbe expensive (Newby et al. 1991) and may furtherimpose mechanical wear. Soiling, especially withhygroscopic compounds (e.g. sulfates), facilitatesthe formation of moisture. Finally, some com-pounds act as catalysts for various reactions —notably oxidation of SO2 and NOx.

Corrosion of metals

In dry air and at normal temperatures oxygen re-acts with most metals and in some cases forms pro-tective layers of oxides, which inhibit furthercorrosion. Aluminum is covered with Al2O3 andiron with a mixture of oxides. Copper is coveredwith an attractive verdigris, which in a pure at-mosphere mainly consists of basic copper carbon-ates. In a polluted atmosphere a less protectinglayer of sulfates is formed.

In humid air the corrosion is generally a morerapid electrochemical process in a moist layer onthe metal surface, where, for example, iron is con-verted to rust (FeOOH), which is peeled or washedoff. Pollutants will dissolve in the moist layer (Fig.15.16) and sulfur dioxide, for example, will be oxi-dized to sulfate:

SO O 2e SO2 2 42+ + Æ- -

Simultaneously, the metal (e.g. iron) will bedissolved:

Ferrous sulfate can be further oxidized to ferric sul-fate and converted to rust. This liberates sulfate,which can act once more. At the same time therust maintains the humid corrosion layer.

The lower levels of SO2 in the industrializedcountries have increased the relative significanceof other atmospheric pollutants. Most importantis NO2. Its corrosive reactions are not fully under-stood and it may not have impacts per se. It ap-pears, however, to increase the effect of sulfurdioxide —possibly by facilitating its oxidation tosulfate (Arroyave & Morcillo 1995).

Stone materials

Stone used as building material and for monu-ments has a wide range of composition, texture,

Fe Fe 2e2Æ ++ -

Atmosphere

Electrolyte

Corrosionproducts

Metal surfaceMetal

O2, CO2, SO2, NO2, etc.

Fig. 15.16 Corrosion of a metal surface is an electrolyticprocess in a thin humid layer. The attack need not beuniform. Often it appears in the form of “pit corrosion,”with more complex reactions in a confined area with atypical diameter of a few millimeters.

416 jes fenger

and structure. Most important is the porosity,ranging from 0.5% for the dense granites and mar-bles to 25% for some limestones and sandstones.The classical pollutant causing degradation is SO2,which is dissolved in the moisture layer to formsulfite and eventually is oxidized to sulfate. Thisattacks calcium carbonate (CaCO3) and forms gyp-sum (CaSO4·2H2O). The process initially results ina weight increase (Fig. 15.17), but in time the gyp-sum is washed off by rain.

In porous materials, notably soft sandstonesused for ornaments, hygroscopic salts can beformed inside the stone. During changes of tem-perature and humidity various transformationsresult in mechanical stress, which enhancesdegradation. This also means that degradation cancontinue after the pollution has been reduced orstopped.

Organic materials

Degradation of organic materials is mainly relatedto ozone, which attacks double bonds in unsatu-rated polymers. This may lead either to chain-scissoring (with loss of tensile strength) or tocross-linking (with loss of elasticity). Thus accel-erated wear of tires in California was an early signof photochemical air pollution.

The exposure to ozone results in fading andcracking of outdoor paints and other coatings, butthe impacts may be difficult to distinguish from di-rect action from sunlight. However, indoor effectshave also been observed. Accelerated laboratoryexposure to a mixture of photochemical oxidantsleads to extensive fading of artists’ colorants, andhumidity enhances the effect (Grosjean et al.1993). It appears that in some museums withoutproper air conditioning, especially in the Tropics,serious fading may result within a few months.

Economic evaluation of material damage

Material damage is better documented than healthimpacts, and experiments are less controversial.Nevertheless, economic evaluations are uncertaindue to a lack of accurate materials inventories. Var-ious estimates of maintenance costs (Mayerhoferet al. 1995; Kucera et al. 1996; Tidblad & Kucera1998) suggest total European damage in the orderof several billion euro per year. Thus the annualsavings from reduced damage to buildings inEurope of implementing the second sulfur pro-tocol (Section 15.5.1) are estimated at about 10billion euro, with about two-thirds in urban areas.

Notwithstanding the uncertainties in suchevaluations, it appears that the savings from re-duced material damage may balance a consider-able part of pollution abatement costs. On theother hand, the impacts can be significantly re-duced or even prevented by a modification or re-placement of materials, e.g. when metals arealloyed or replaced by plastic.

A special problem is posed by damage to irre-placeable cultural monuments and art, the reasonbeing partly that the value is often related to a thinlayer of decoration and not to mechanical strength.Attempts to evaluate the benefit derived in theform of pleasure from such objects (contingent val-uation) suggest that the economic loss of damage isin the same order of magnitude as that of damage totrivial materials.

15.4.3 Urban ecosystems

The natural environment in urban areas is of im-

Humid climate (TOW=0.6)Moderate climate (TOW=0.4)Dry climate (TOW=0.2)

14

12

10

8

6

4

2

010 20 30 40 50 60 70 80 90 100

SO2 (mg m–3)

Wei

ght

incr

ease

(g m

–2yr

–1)

Fig. 15.17 Weight increase of sandstone calculated froma dose–response relation showing the effect of differentclimates on the deterioration rate. TOW is the so-calledtime of wetness, here defined the fraction of time, whenthe temperature is above 0°C and the relative humidityabove 80%. (From Tidblad & Kucera 1998.)


portance for various reasons (Ashmore 1998, andreferences therein). First, it has utilitarian value.Many city dwellers grow food for their consump-tion, and it is estimated that a total of 800 millionpeople are engaged in urban agriculture —mainlyin Asia, Africa, and Latin America. Commercialproduction of high-value crops is normally foundin the outskirts, but in densely populated regionslike the Netherlands more intensive agricultureactivity is incorporated into urban planning.

Second, vegetation has aesthetic value and treesmay improve the urban climate by offering shadeand cooling by transpiration, and by influencingdeposition. It appears that urban trees generallyreduce ozone concentrations in cities, but tend toincrease average ozone concentrations in the sur-rounding area (Novak et al. 2000). It has even beenargued that urban vegetation may act as a sink formany pollutants. On the other hand, trees tend toreduce the local dispersion of pollutants.

Third, sensitive elements may provide usefulbio-indicators of trends in air quality. The diversi-ty of lichen has in many countries been used to in-dicate levels of SO2. Lichens have also been usedfor mapping deposition of metals, which they ac-cumulate effectively.

Early observations testify to significant damageto urban vegetation. Although urban air qualityhas improved considerably in recent years there isevidence of continuing impacts (Ashmore et al.1998).

15.4.4 Visibility

The most visible air pollutant is black smoke,which acts as a precursor to fogs (an importantingredient in early English detective novels). Al-though the London “Pea souper” and other spec-tacular events (Fig. 15.18) are things of the past,reduction of visibility may still be a nuisance.

It is pleasant when the air is clear and it is ad-vantageous in many ways, including road safety. Areduction in visibility is therefore not only unsa-vory, it may also be dangerous. What determineswhether we can see an object of a reasonable size isnot so much its radiance as its contrast —normallydefined as the relative difference between the lu-

minosity of the object and that of the background.When light passes through the atmosphere it is ab-sorbed and scattered by molecules and particles,and consequently the contrast is gradually re-duced. For contrasts below 0.02–0.05 the objectbecomes for most people indistinguishable fromthe background, and the corresponding distance istermed the visibility.

In practice visibility is determined by light scat-tering on particles, and there are different pro-cesses, which depend upon the ratio between the wavelength and the particle size. Thus notonly the contrast, but also the color mix, of sight ischanged. For relatively large particles like fog thescattering is largely independent of the wave-length and results in a diffuse white light, whichmay reduce visibility to nearly zero.

When the relative humidity becomes so largethat particles adsorp water their size will increasedrastically and so will the scattering —especiallyfor water soluble compounds like sulfates andnitrates.

The number of particles in urban air is typicallyhighest in the morning and afternoon, not only be-cause of rush hours, but also because of a low-lyingmixing layer. At the same time the humidity maybe high and thus the visibility low (Fig. 15.19).

In clean dry atmospheres, as in some naturalparks, visibility may be as far as 200km. In

15

10

5

011:00 12:00 13:00 14:00 15:00 16:00

GMT

Light

inte

nsi

ty (

kilo

lux)

Fig. 15.18 The light intensity in London on the “DarkDay,” January 16, 1955. Here the smoke concentrationat ground level was not particular high, but a higher layer practically blocked all sunlight. (From Brimblecombe 1987.)

418 jes fenger

European cities (without direct fog) visibility maybe down to 2km. It is thus rarely a practical, onlyan aesthetic, problem. In the case of more seriousepisodes pollution may build up over several days(Section 15.3.2).

15.5 MEANS OF MITIGATION

As indicated in Fig. 15.1 the impacts of urban airpollution can be mitigated at all points of the “pol-lution chain.” Some possibilities have alreadybeen mentioned. The classic solution, with disper-sion from high stacks (Fig. 15.20), tends to onlytransfer the problem. With the growth and eventu-al merging of urban areas this option is only ac-ceptable in combination with advanced flue gascleaning.

The problem with increasing traffic and itsemissions can in principle be attacked in twoways: on the technical level by reductions of theindividual emissions; and on the economic andplanning level by a reduction of activities, eitherby reducing the need for transport as such (promot-ing walking and bicycling) or by promoting publictransport. Some solutions based on traffic plan-ning, however, are of debatable value. They im-prove the air quality in city centers, but transferproblems to the outskirts and may even increasetotal activity.

15.5.1 Legislation and technical solutions

Past experiences have with depressing clarityshown that existing technical possibilities andrecommended management practices will not beimplemented unless legally or economically en-forced. Air quality expressed as pollutant concen-trations is controlled by limit values. With thepossible exception of ozone limit values, they areonly relevant for pollution levels in urban or indus-trial areas. The scientific foundation is experi-ments on humans or animals and epidemiologicalinvestigations. The results are evaluated by theWHO and expressed in the form of guidelines(WHO 2000) that are subsequently used as thebasis for legally binding limit values.

Most countries have established such limitvalues for the major air pollutants and use, inaddition, guideline values for a series of other com-pounds. Most important are the limits in the EUand the USA, which in many cases have served asmodels for other regions. US emission standards,especially for motor vehicles (Faiz et al. 1996),have been used in this way.

20

15

10

5

0

0.65

0.60

0.55

0.50

0.45

0.40

0.35

0.30

0.258 9 10 11 12 13 14 15 16

StandardvisibilityHumidity

Rel

ati

ve h

um

idit

y

Vis

ibili

ty (

km)

Time of day (h)

Fig. 15.19 Variation of visibility as a function ofhumidity in Vienna on a day with stagnant air. Note thatthe humidity and the visibility vary inversely. (FromHorvath 1998.)

Fig. 15.20 The policy of high stacksonly transferred the problems. (Anearly cartoon from Euroforum.)


Air quality in the EU

In the EU the setting of limit values is a multistepprocess (Edwards 1998) involving a system of EUdirectives, the first being adapted in 1980. Since1996 a framework directive has provided a basicstructure, and daughter directives lay down limitvalues and proscribe dates for attainment, meth-ods for measurements, etc., which are mandatorythroughout the EU. These directives are subse-quently ratified in the individual member states inthe form of national legislation (Edwards 1998).This system has so far covered sulfur dioxide, par-ticulate matter, nitrogen dioxide, benzene, carbonmonoxide, and lead. Threshold values for ozone forthe giving of information and warnings to the pub-lic are also regulated by EU directives. Other pollu-tants for which legislation is considered importantinclude polyaromatic hydrocarbons, cadmium, arsenic, nickel, and mercury.

Vehicle emissions in the EU

The most direct means of improving air quality isof course through regulation of emissions. The EUlegislation on vehicle emissions and fuel qualitystandards has evolved greatly since the first direc-tive in 1970. The early legislation had the dual pur-pose of reducing pollution and avoiding barriers totrade due to different standards in different mem-ber states. It is now aiming at meeting air qualitytargets. The “Auto Oil I Programme,” undertakenby the European Commission in conjunction withindustry (European Commission 1996a), has set uptargets for a series of traffic-related pollutants andassessed different technologies and fuel qualitystandards. The target for nitrogen dioxide was infull compliance with the new WHO guideline of200mgm-3 as a maximum one-hour average.

By means of models with simplified chemicalreactions, the impacts, compared to 1990, on theair quality in seven representative European cities(Athens, Cologne, London, Lyon, Madrid, Milan,and The Hague) have been evaluated (EuropeanCommission 1996b). Already agreed measures(e.g. the introduction of three-way catalytic con-verters) were expected to reduce pollution from ve-

hicles by 40–50% in 2010 compared to 1990. TheAuto Oil Programme will increase the reduction to70% —even for the expected higher traffic volume(Fig. 15.21).

It appeared that the objectives for carbonmonoxide (10mgm-3, one hour maximum), ben-zene (10mgm-3 annual mean), and particulate mat-ter (50mgm-3 24-hour average) would be met by2010, and the NO2 objective should be met in mostof the Union in 2010. In some cities, like Athens,however, further action would be needed. A morestringent target value for benzene of 2.5mgm-3,preferred by several member states, would be ex-ceeded in all investigated cities except The Hague.

In an ongoing second program nontechnicalmeasures for local use in areas with high pollutionlevels, such as road pricing, traffic management,and scrapping schemes, will be evaluated. Further,the resulting reductions in pollution levels will be calculated in more detail. An example from aDanish site is shown in Fig. 15.22.

Stationary sources in the EU

EU legislation on industrial air pollution has al-

1990 1995 2010 (without) 2010 (with)120

100

80

60

40

20

0NOx

urbanCO

urbanBenzeneurban

Particlesurban

NOxtotal

VOCtotal

Perc

enta

ge

Fig. 15.21 Expected impacts of the European Auto OilProgramme on emission of air pollutants in theEuropean Union and specifically in urban areas. Thediagram shows calculated emissions in 1990 and 1995and projected emissions in 2010 without and with theprogramme in force. (Based on EC Commission 1996a.)

420 jes fenger

ways taken air quality limit values into accountand required the operators to use “best availabletechnology, not entailing excessive costs.” ACouncil Directive from 1996 on integrated pollu-tion prevention and control (IPPC) has the purposeof reducing pollution to the environment as awhole, avoiding transfer from one medium to an-other (Edwards 1998).

The United States Clean Air Act

In the USA the first federal air pollution legislationwas enacted in 1955, and in 1970 the administra-tion was transferred to the new US EnvironmentalProtection Agency (EPA). The Clean Air Act wasfirst passed by Congress in 1967 as the Air QualityAct, which was amended in 1990. Under the act,the EPA set national air quality standards(NAAQS) for six pollutants: sulfur dioxide, nitro-gen dioxide, carbon monoxide, particulate matter,and lead.

The standards are reviewed regularly, and new proposals were put forward in 1997. For ozonethe new EPA standard allows no more than 0.08p.p.m. (160mgm-3) as an eight-hour average. For

particles a separate standard is set for PM2.5 (Brown1997).

Long-range transport and urban air pollution

The United Nations Economic Commission forEurope (UNECE), comprising all European coun-tries and Canada and the USA, has been an impor-tant forum for east–west discussions of airpollution. It was also in the ECE that the so-calledGeneva Convention on long-range, transboundaryair pollution was established and signed in 1979.After it had been ratified by a sufficient number ofmember states it came into force in 1983. A seriesof related protocols set targets for reductions of na-tional emissions of sulfur dioxide, nitrogen diox-ide, and volatile organic compounds.

These protocols are all aimed at protecting natural systems. A new multipollutant–multi-effect “Gothenburg Protocol” comprehensivelyaddresses acidification, eutrofication, and photo-chemical air pollution. However, since the largestpart of the relevant emissions occur in urban areas,the necessary reductions have direct impacts onurban air quality. Thus the decreases in levels ofsulfur dioxide in the 1980s are related to the use ofnatural gas, low-sulfur fuel oil, and desulfurization(or alternative sources of energy), necessary tocomply with international agreements. Attemptsto reduce ecological impacts of large-scale photo-chemical air pollution will directly influenceurban ozone levels —especially in the north of Europe.

In other parts of the world, notably in east Asia,unregulated local air pollution emissions have in-creasing regional and transboundary impacts.

15.5.2 Economic incentives

Attempts to reduce urban driving by various typesof economic incentives have had some success,but they are often opposed by trade. It must also beconsidered that the placing of restrictions in citiesmay promote the growth of big shopping centers,hotels, and office buildings outside the cities,where they can offer free parking space and otherfacilities, often resulting in an increase in total

1995 2000 2005 2010 2015 2020

60

50

40

30

20

10

0

NO

2 (

mgm

–3)

EU-limit volue

Fig. 15.22 The yearly average of NO2 at Jagtvej inCopenhagen, a street with high buildings and a presenttraffic of about 22,000 per day, which is not expected torise. For 1995 the value is based on measurements. Latervalues are calculated with a street pollution modelcoupled to a simple model for urban backgroundpollution (Berkowicz 2000). The EU limit value of 40mgm-3 will easily be met.


traffic volume (OECD 1995b). In such cases im-provement in urban air quality is paid for withmore pollution on a larger geographic scale.

Taxes

Taxes can be imposed on the purchase of a vehicleor as an annual tax related to vehicle weight, ener-gy consumption, or emissions. Generally, the im-pact on traffic pollution is a reduction, but it mustbe taken into account that purchase taxes discour-age the replacement of an outdated automobilefleet. Fuel taxes will in the long run lead to more effective driving and increase the demand forsmaller and more efficient automobiles. The totaleffect on energy consumption and emissions ismuch higher (and more logical) than that of anyother form of taxation.

The basic problem with taxes is that the real po-litical purpose often appears to be a means to pro-vide additional revenue rather than protection ofthe environment. In addition, attempts to reducetransport contrast with the general belief that mo-bility in society is a prerequisite for economicgrowth.

Parking and driving restrictions and fees

More rational types of economic incentive areaimed at reducing traffic in sensitive areas. Dri-ving in city centers can thus be reduced by restrict-ing the number of parking places or by increasingthe fees for their use.

Pricing of the infrastructure in the form of tollroads is used in several areas. Typically, a fee is paidfor the use of highways, but pay stations can alsoform a ring around a town to tax all traffic in theurban area.

A more advanced, in principle more fair, but notyet fully developed, system is road pricing. Herethe position of each automobile is determined bysatellite or by coils in the road network and the in-formation on movements is stored electronicallyin the automobile. This allows detailed taxing ac-cording to type of vehicle, different zones, and timeof the day, etc., and thus encourages economicaldriving. The general objection to the system is the

vision of a “big brother society” where all move-ments of all people are registered. This, however,appears to be only a technical problem.

15.5.3 City planning

The result of a choice between public and individ-ual transportation depends upon the proximity ofaccess to the public system. In an industrializedcountry a distance of more than 1km often meansthe use of an automobile (Fig. 15.23).

The impacts of urban air pollution can thereforebe mitigated by constructive city planning. Thecomplete separation of industry and habitation,originally envisaged as an environmental im-provement and a reasonable solution in a societywith heavily polluting industries, is now outdatedand often only leads to increased commuting traf-fic and congestion. The ideal now is integratedland use, which minimizes transport and thustotal urban emissions. Open spaces and parks canbe used to improve environmental quality.

In existing cities the possibilities for restructur-ing are limited, but the construction of ring roads,which lead part of the traffic around the city cen-ter, is one of the options (OECD 1995b).

In the industrialized world few cities and urban

80

60

40

20

0

Central area Suburbangood public

traffic service

More than 1kmto public

transport terminal

Perc

enta

ge

Car

Public

Walker/bicycle

Fig. 15.23 Distance and choice of mean oftransportation in the Copenhagen area. (From Nielsen1997.)

422 jes fenger

areas can be constructed from scratch, but whenpossible new concepts of integrating urban plan-ning, building design, and supply of renewable en-ergy should be applied. The climate of the city isalso important, and the influence of buildings and street canyons on solar radiation, shade, andwind pattern should be taken into account (Bitan1992).

In this planning, which to a large extent is plan-ning of traffic, it must be realized that air pollutionis not the only environmental impact, and proba-bly not even the most important. It is estimated(European Commission 1995) that on average theexternal costs of air pollution (not including thegreenhouse effect) from transport in the EUamount to 0.4% of the gross national product,compared to 0.2% from noise, 1.5% from acci-dents, and 2.0% from congestion.

Figure 15.24 shows as an example of a Danishsuburb west of Copenhagen, where the urban den-sity increases close to a railway station. In this areapublic facilities such as a shopping center andschools are located within walking distance. Theunintended side-effect of such a solution, wherenearly everything is built at one time, is often a ho-mogeneous and uninspiring style. (Apparentlyvery few architects live in such developments.)

Even for a fixed volume of traffic in a giveninfrastructure the pollution from traffic can be re-duced. The emissions per kilometer from a specif-ic automobile depend upon the driving pattern andincrease drastically with “stop and start” drivingin congested urban areas. Here average speed ap-pears to be a good measure of the number of accel-erations and stops and thus low speeds result inhigh specific emissions. Smooth traffic (securedwith speed limits and green waves), which increas-es the average speed from 20 to 40km in mainflows, may thus reduce the emissions by 30–40%(Fig. 15.25).

15.6 CASE STUDIES

The complex interplay between human activities,technical and legislative development, and naturalparameters gives rise to completely different

pollution patterns in different cities of the world.However, there are some common features forcities in the industrialized world (mainly exempli-fied by western Europe), the developing world(mainly east Asia), and economies in transition(eastern Europe), respectively.

15.6.1 Comprehensive records of urban air quality

In most of the industrialized world urban air pollu-tion is now monitored routinely. Since 1974 WHOand UNEP have, within the “Global EnvironmentMonitoring System,” collaborated on a project tomonitor urban air quality, the so-called GEMS/AIR (UNEP 1991; WHO/UNEP 1992; GEMS/AIR1996; and a series of related reports). Concen-trations of air pollutants in selected countries arealso reported yearly by the OECD (1999). A com-prehensive presentation of urban air pollution in Europe, based on data from 79 cities in 32 countries (Richter & Williams 1998), has recentlybeen published by the European EnvironmentAgency.

These and similar data give an indication oftrends in ambient air quality at the national leveland in selected cities. Often, however, the data arebased on only a few monitoring stations placed atcritical sites, and thus represent micro-environ-ments. It should also be taken into account thatthe coverage of stations varies from country tocountry (Larssen & Hagen 1997), and that averagevalues can therefore be differently biased.

Air pollution in the developing countries and insome countries with economies in transition isnot documented in detail and longer time series are very rare. In most cases a general trend in airquality can only be estimated on the basis of dubi-ous emission inventories. Data presented in theopen literature are seldom up to date and normallyconcern specific cities, which may not be fully rep-resentative. In recent years many governmentaland private institutions from the industrializedcountries have acted as consultants in developingcountries or performed investigations, but not all the efforts have been reported in the openliterature.


Railway station

School and shopping center

100m

Fig.15.24 Albertslund, west of Copenhagen, is a typical example of an urban development aimed at public transport.(From Gaardmand 1993.)

424 jes fenger

15.6.2 General pollution development

Seen over longer periods, pollution in major citiestends to increase during the build-up phase, passthrough a maximum and then become reduced, asabatement strategies are developed (Mage et al.1996; Goklany 1995). Depending upon the time of initiation of emission control the stabilizationand subsequent improvement of the air qualitymay occur sooner or later in development (WHO/UNEP 1992).

In the industrialized Western world urban airpollution is in some respects in the last stage, witheffectively reduced levels of sulfur dioxide andsoot. In recent decades, however, increasing traffichas switched the attention to nitrogen oxides, or-ganic compounds, and small particles. In somecities photochemical air pollution is an importanturban problem, but in the northern part of Europeit is a large-scale phenomenon, with ozone levelsin urban streets normally being lower than in ruralareas. Cities in eastern Europe have been (and inmany cases still are) heavily polluted. After the re-cent political upheaval, followed by a temporaryrecession and subsequent introduction of newtechnologies, the situation appears to be improv-ing. However, the rising number of private auto-

mobiles is an emerging problem. In most develop-ing countries the rapid urbanization has so far re-sulted in uncontrolled growth and deterioratingenvironment. Air pollution levels are still risingon many fronts.

In accordance with this development, the envi-ronmental quality in a given country generally de-pends upon the average income of the inhabitants(Shafik 1994). The availability of safe water and ad-equate sanitation increases with income, and sodoes the amount of municipal waste per capita. Airpollution, however, appears initially to increasewith income up to a point and then to decrease (Fig.15.26). Based on more recent data, Grossman andKrueger (1995) estimate that the turning point fordifferent pollutants varies, but in most casescomes before a country reaches a per capita incomeof US$8000. Although this suggests that globalemissions will decrease in the very long term, con-tinued rapid growth over the next several decadesmust be expected (Selden & Song 1994). Emissionof CO2, the (so far) unavoidable end-product ofmost energy production still increases in all cases.

Fully in line with this typical development, theWorld Commission on Environment and Develop-ment in its report “Our Common Future” (1987)conceives technological development and risingstandards of living as a prerequisite for environ-mental improvement. Or as Bertholt Brecht hasput it “Erst kommt das Fressen, dann kommt dieMoral” (mistranslated into English: “First devel-opment and only later pollution control”).

15.6.3 Industrialized (OECD) countries

Europe

Europe is a highly urbanized continent with morethan 70% of the population living in cities. Popula-tion changes are in most countries modest andpartly due to refugees and migration from east towest.

The large resources of coal were the primarysource of energy during the Industrial Revolution,but in recent decades oil and gas have been found inthe North Sea and in Russia and have been trans-

Em

issi

on (

gkm

–1)

4

3

2

1

00 10 20 30 40 50 60 70

HC

CO

NOx

Speed (kmh–1)

Fig. 15.25 Average speed and emissions for a Danishpassenger car around 1990. Although the technology hasbeen continuously improved the tendency is obvious.(From Krawack 1991.)


ported through pipelines to other regions. In west-ern urban areas the consumption of coal in smallunits, e.g. used for domestic heating, has beenreplaced with less polluting fuels, resulting inreduced emissions of sulfur dioxide and particles.

The total emission of sulfur dioxide steadily in-creased from about 5Mt in 1880 to a maximum ofnearly 60Mt in the 1970s, only interrupted by theSecond World War. It peaked in the mid-1970s, buthas now been reduced to less than half. For thetraffic-related pollutants nitrogen oxides, carbonmonoxide, and VOCs, an increase has only recent-ly been reversed (10–15%) by the introduction ofthree-way catalytic converters (TWC) (EMEP/MSC-W 1998, and related earlier and later reports).

European road traffic currently accounts for the largest part of total CO emissions, more thanhalf of the NOx-emissions, and a third of the VOCemissions, but only a minor part of sulfurdioxide —in western and northern cities about 5%and in southern cities, where diesel oil has a highersulfur content, 14%.

It is estimated that in 2010 the CO, VOC, andNOx pollution from traffic within the EU will bereduced drastically despite the expected trafficincrease (Figs 15.21 and 15.22).

The European cities differ in various ways,which influence the relations between emissionsand resulting pollution levels. Western Europe is influenced by the predominant westerly wind,bringing moist air from the sea, a climate that alsofavors long-range transport. In the northern part ofEurope the small amount of sunlight favors per-sistent inversions with poor dispersion condi-tions. In central and eastern Europe high pressure,with air stagnation and accumulation of local pol-lution, is frequent. During the summer the climatein the Mediterranean Region likewise favors accu-mulation of local emissions, whereas during thewinter large-scale wind systems are more fre-quent. The formation of photochemical oxidantsdepends upon sunlight, which in combinationwith poor dispersion conditions results in frequentepisodes during summer.

The most extensive comprehensive treatmentof Europe’s environment up to 1992 was given inthe so-called Dobris Assessment (Stanners &

100 1000 10,000 100,000

100 1000 10,000 100,000

100 1000 10,000 100,000

0

10

20

30

40 CO2

90

45

0

SO2

1972

1986

1972

1986

Perc

enta

ge

of

popula

tion

120

80

40

0

mg m

–3

Met

ric

tonne

per

per

son p

er y

ear

TSP

Fig. 15.26 Urban concentrations of SO2, particulatematter, and CO2 as a function of per capita income(Shafik 1994). The curves only show general tendencies,there are marked deviations. Thus in the early 1980sKuwait had both some of the highest incomes andhighest pollution levels. (From Smith 1988.)

426 jes fenger

Bourdeau 1995) from the European EnvironmentAgency. It has since been updated in a second as-sessment (EEA 1998). The general air quality in European cities has improved in recent decades —often despite an increase in population density andstandard of living —but air pollution is still consid-ered a top priority environmental problem withboth urban and large-scale impacts. Its special as-pects are treated in more detail in several reportsfrom the agency (e.g. Jol & Kielland 1997).

Smoke. The drastic improvements in urban airquality in the 1960–70s, partly brought about by achange from coal to less polluting fuels for domes-tic heating, partly by the closing down of pollutingindustry, has also resulted in a marked reductionin incidence of fog (e.g. Eggleston et al. 1992), butthe lack of chemical analysis and size fractiona-tion preclude more than a qualitative evaluation ofthe health impacts.

Sulfur. In many western European cities severepollution with sulfur dioxide is a thing of the past.In Copenhagen the yearly average of SO2 concen-tration was about 80mgm-3 around 1970 and dur-ing the winter about 120mgm-3. Today it is wellbelow the WHO guideline of 50mgm-3. In provin-cial towns it is even lower (Fig. 15.27). The levels inprovincial cities are not much higher than those atrural sites, indicating that most of the sulfur diox-ide is due to long-range transport.

In accordance with this a more detailed analysis(Kemp & Palmgren 1999) has shown that there arestill significant peak values (95th and 98th per-centiles), identified as being due to long-rangetransport from central and eastern Europe. Consis-tent with this, levels of particulate sulfur (sulfate)show only a slight downward trend up to 1996. In1997 levels for all types of sulfur pollution werevery low, but so far it has not been establishedwhether this was due to reduced emissions or spe-cial meteorological conditions.

This development has had several causes. Animportant aspect is the Geneva Convention ontransboundary air pollution, which resulted in re-duction of total emissions, but a widespread tran-sition from individual to district space heatingproduced in large units with high stacks (often ascombined heat and power production) has played arole. A further factor is that the oil crisis in theearly 1970s resulted in better insulation of build-ings. In Denmark the consumption of energy usedfor space heating was reduced in the period1972–82 by about 30% despite an increase in areaheated of about 20% (ENS 1997).

In 1990, 10 European cities observed ex-ceedances of the long-term WHO-AQG for SO2 of50mgm-3; in 1995 only Katowice and Istanbul did.The short-term guideline of 125mgm-3 daily aver-age is, however, still exceeded for a few days peryear in many countries.

88 91 94 97 88 91 94 97 88 91 94 97 91 94 97

25

20

15

10

5

0

SO2 (

mgm

–3)m

edia

n

Copenhagen Odense Aalborg Lille Valby

Fig. 15.27 Trends for annualmedians based on hourly averageconcentrations of SO2 measured inthe center of the Danish capitalCopenhagen (1.8 millioninhabitants), in three majorprovincial cities, and at a ruralstation (Lille Valby). (From Kemp &Palmgren 1999.)


Nitrogen oxides. In recent years urban air pollu-tion nitrogen dioxide has shown a downward trendin most European cities, although the short-termWHO guideline (corresponding to 200mgm-3 asmaximum hourly value) is exceeded in cities (EEA1998).

The situation is complicated by chemical reac-tions in the urban atmosphere. Since the introduc-tion of catalytic converters on all Danish gasolineautomobiles registered after 1990 there has been amarked reduction in the levels of NO in Danishcities. Levels of NO2, however, have been reducedmuch less (the median in Copenhagen by about20%), reflecting the point that emissions of NO arenot the limiting factor, but the available O3 (Kempet al. 1998).

Carbon monoxide. Urban concentrations of car-bon monoxide have likewise been reduced since1990, although exceedances of the eight-hourWHO guideline have been reported from manycities (EEA 1998).

Lead. In countries that have reduced or eliminat-ed lead in gasoline the lead levels have substanti-ally declined (EEA 1998), especially in countrieswith few or no lead-emitting industries. Since automobile exhaust is emitted at low height andoften in street canyons there has been a close corre-spondence with the lead concentrations in urbanair. In Denmark, where lead from gasoline in 1977accounted for 90% of the national lead emissions,the problem has virtually disappeared. Remaininglow lead concentrations of about 20ngm-3 are es-sentially due to long-range transport (Fig. 15.28).

In the early 1980s, 5% of Europe’s urban popula-tion in cities with reported lead levels were ex-posed to more than the WHO guideline of 0.5mgm-3. At the end of the decade levels above theguideline value were no longer reported from thewestern countries.

Volatile organic compounds. Recently published(EEA 1998) levels of benzene range from a few tomore than 50mgm-3, with the highest values nor-mally found near streets with high traffic volume.Of the reporting cities only Antwerp did not ob-

serve exceedances of the WHO-AQG guideline of2.5mgm-3 as a yearly average.

In a study of personal benzene exposure in a se-ries of European cities (see Sections 15.2.2 and15.4.1) it was shown that the levels decreased fromnorth to south, probably reflecting a decrease indispersion (Fig. 15.29).

In the 1960s the annual average concentrationof BaP was above 100ngm-3 in several Europeancities (WHO 2000). In most developed countriesimproved combustion technology, a change offuels, and catalytic converters in motor vehicleshave reduced the urban levels to 1–10ngm-3.However, urban air pollution by potentiallycarcinogenic species is still not satisfactorilyunderstood.

Ozone. European ozone levels appear to have in-creased from about 20mgm-3 around 1900 to aboutdouble now, with the most rapid rise between 1950and 1970, concurrent with the rise in emissions ofprimary precursors (Volz & Kley 1988). In the out-skirts of Paris the early ozone levels were about 20mgm-3, but in the center only 3–4mgm-3. Sincethere was no nitric oxide from motor vehicles it

Lead e

mis

sion (

tyr–

1)

1200

1000

800

600

400

200

01970 1975 1980 1985 1990 1995

1.2

1.0

0.8

0.6

0.4

0.2

0

Emission

Lead in air

Lead inpetrol

Lead in a

ir (

μg m

–3)

Lead in p

etro

l (g

l–

1)

Fig. 15.28 Annual average values for total Danish leademissions 1969–93, lead pollution in Copenhagen since1976, and the average lead content in gasoline sold inDenmark. (From Jensen & Fenger 1994.) The dates oftightening of restrictions on lead content are indicatedwith bars. Lead concentrations for recent years can befound in Kemp et al. (1998, 1999). In 1998 they werebelow 20ngm-3.

428 jes fenger

is assumed that O3 was reduced by SO2 or NH3(Anfossi & Sandroni 1997).

Summer smog with high ozone concentrationsoccurs in many European countries. As an urbanphenomenon it is most serious in Athens andBarcelona, with concentrations up to 400mgm-3,but Frankfurt, Krakow, Milan, Prague, andStuttgart are also affected. Generally the presentEuropean ozone concentrations decrease fromsoutheast to northwest.

An EU ozone directive contains a thresholdvalue for giving information to the population of180mgm-3 and for warning of 360mgm-3. In theNordic countries it is seldom necessary to givemandatory ozone information and never mandato-ry warning. Implementation of the VOC protocolis expected to result in a 40–60% reduction in highpeak values and 1–4% in annual average O3 con-centrations (EEA 1998).

General trends. According to the OECD (1998)the typical reduction of sulfur dioxide in WesternEuropean cities in the period 1988–93 was nearly40% for traffic sites and about 20% for residentialareas. For NO2 the reductions were 10 and 12%respectively. Levels of CO were in generalunchanged.

The urban population exposure has decreasedcorrespondingly. In a WHO (1995) study, a series ofdata for the period 1976–90 from European urbanareas with populations above 50,000 were pooledin two groups: up to and after 1985. Notwithstand-ing the limited and not fully representative data itappears that in western Europe up to 1985, 58%were living in areas with annual mean concentra-tions above the WHO air quality guideline of 50mgm-3. After 1985 the fraction has been reduced to14%. Similar, but smaller, reductions were ob-served for black smoke and SPM. Only for NO2 wasa slight increase observed (WHO 1995). The num-ber of peak values has also decreased, and the totalpopulation experiencing episodes exceeding 250mgm-3 SO2 decreased during the 1980s from 71to 33%.

North America

The two countries in North America, the USA andCanada, are among the wealthiest in the worldwith respect to both natural resources and produc-tion. This has previously led to serious urbanpollution especially in the USA.

Many cities in the early industrialized USAwere, like those in Europe, characterized by heavysmoke, and have subsequently gone through thetypical development process. As an example(Davidson 1979), Pittsburgh already had pol-lution problems in 1804, and they by and large in-creased until the first effective smoke control inthe late 1940s after the city experienced severepollution due to the Second World War steelproduction.

As a general measure of the development in airpollution in the period 1970–93, the overall emis-sions of carbon monoxide, volatile organic com-pounds, and particulate matter have been reducedby 24, 24, and 78% respectively. A decline in emis-sions of nitrogen oxides from vehicles has been off-set by increased electricity generation, giving anoverall increase of 14%, but since the emissionsare generally not in urban areas, this has not pre-vented a reduction of NO2 levels in cities. Lead asan air pollutant has virtually disappeared, buttoxic chemicals are still a problem.

30

25

15

10

5

0

0.0 1.0 2.0 3.0 4.0 5.0 6.0

20

Padua

Athens

Antwerp

Copenhagen

Ben

zene

(μg

m–

3)

Wind speeds (m s–1)

Fig. 15.29 Benzene levels as a function of wind speed ina series of European cities. The same pattern is observedwhen the levels are plotted against the latitude of thecity. (From Cocheo et al. 2000.)


Los Angeles still has the largest ozone problemin the USA. In 1990 the highest one-hour averagewas 660mgm-3. Ozone concentrations showed asignificant decrease of 30% from 1988 to 1993 inurban residential areas in the USA both as an aver-age and in the most polluted cities. Many countieswill find it hard to comply with the new ozonestandard of 0.08 p.p.m. (Brown 1997).

According to the OECD (1998) the trends in airquality in city areas in the USA in the period1988–93 were decreases of the order of 23% forSO2, 11% for NO2, and 22% for CO.

In Canada emissions of sulfur dioxide and par-ticulate matter have been reduced significantlysince the early 1970s, and lead has virtually disap-peared. However, some central Canadian citiesstill experience unacceptable air quality, with highlevels of ozone and particulate matter, especiallyduring the summer.

Japan

Japan is in many respects not typical of east Asia,but more like the western countries. According tothe OECD (1998) the trends in air quality in Japan-ese city areas in the period 1988–93 were decreasesof 20% for SO2 and 26% for NO2, whereas CO waslargely unchanged.

The capital of Japan, Tokyo, is an encouragingexample of an industrial megacity where air pollu-tion is controlled. In the 1960s it was heavily pol-luted due to coal combustion and insufficientemission controls. Concentrations peaked in thelate 1960s with annual mean values for SO2 andTSP up to200 and 400mgm-3 respectively (Komeijiet al. 1990). By switching the major fuel consump-tion from coal to oil and installing of dust collec-tors the annual mean values of SO2 and TSP werebrought well below WHO guidelines in the 1980s(WHO/UNEP 1992), and as a spectacular examplethe reduction of particulate emissions in the 1970s has resulted in a corresponding increase invisibility (Fig. 15.30).

As in the Western world, stricter control of ve-hicle emissions has been counteracted by a growthin traffic and the levels of traffic-related pollutionhave merely been stabilized. However, ozone lev-

els have generally been halved since 1970, whenthe yearly average was about 80mgm-3.

15.6.4 Economies in transition

In Eastern Europe solid fuels are still used in pri-vate houses and industry. Many buildings arebadly insulated with large potential savings. Com-bined heat and power production is limited, partlydue to problems with financing and management.

In 1990 the energy intensity of economies incentral and eastern Europe (CEE) was about threetimes higher than that in western Europe; per unitof GDP emissions of NOx and SO2 were more thanfour times higher; and emissions of particles andVOCs were considerably higher (Bollen et al. 1996;UNEP 1997). SO2 emissions were highest in thenorthern part of the region (Poland), which de-pends heavily on indigenous coal and lignite(Adamson et al. 1996), and lower in the southernCEE, where oil and gas are available.

In general it appears that industrial “hot spots”have shifted from western Europe to the east andsoutheast, where heavy industry, the use of low-quality fuels, and outdated production technolo-gies have resulted in high emission levels.

Nevertheless, emissions have been reduced insome areas, partly as a result of German reunifica-tion in 1990 and the collapse of the Soviet Union in

The Tokyo tower

Mt Fuji

1965 1970 1975 1980 1985 1990 1995

180

160

140

120

100

80

60

40

20

0

Annual vi

sible

day

s

Fig. 15.30 The concentration of particulate pollution in Tokyo (right scale) and the days per year when theTokyo Tower and Mount Fuji were visible from asuburban site. (From Kurashige & Miyashita 1998.)

430 jes fenger

1991. In 1990–1 distribution problems, ethnic con-flicts, and organized crime in combination withoutdated technology resulted in a drop in the Sovi-et Russian gross domestic product of 10–15%. Andin the period 1990–6 NOx emissions dropped byone-third and SO2 by one-quarter.

A typical Polish example is Katowice (GEMS/AIR 1996), where since the late 1980s many oldermore polluting industrial plants have been closeddown. In the period 1988–92 this —in combinationwith the introduction of better emissioncontrols —resulted in a reduction of emissions ofnitrogen oxides and particulate matter of 22 and41% respectively. In the same period the air quali-ty was improved by 30 and 44% for SO2 and NO2re-spectively. Such reductions are also reflected inthe investigations of population exposure carriedout by the WHO (1995).

Further substantial reductions in emissions arepossible, even if only current western Europeanpractices are applied (Bollen et al. 1996), but it is aserious problem for CEE to raise funds for econom-ic and technological growth in connection with atransition to a market economy. For some cities,e.g. Krakow, the most effective strategy to improveair quality was found to be a ban on the use of coal —possibly limited to the town center(Adamson et al. 1996).

The southeastern urban areas have a partly out-dated automobile fleet and are decades behind inthe organization of road traffic. Only recently haveefforts been put into the construction of ring high-ways (e.g. in Budapest and Prague) to reduce unnec-essary crossing of the city center. Lead pollution isstill found in eastern Europe, notably in Romaniaand Bulgaria near large uncontrolled metal indus-tries (WHO 1995). The worst environmental prob-lems are probably in the Balkans (Brown 1999),where they are aggravated by armed ethnic con-flicts bordering on civil war.

Despite the general improvement in Europeanair quality the WHO short-term air quality guide-lines for SO2 and TSP are often violated duringwintertime smog, with the highest exceedancesobserved in central European cities. A positive de-velopment in the entire region can be hoped forwith the extension of the EU.

15.6.5 Developing countries

Latin America

Latin America and the Caribbean is the mosturbanized region in the developing world, with arapidly increasing vehicle fleet, which is the domi-nant source of air pollution (Onursal & Gautam1997). In Mexico City it accounts for 99% of theCO, 54% of the hydrocarbons, and 70% of NOx.Leaded gasoline is still permitted in most coun-tries and even in 1995 constituted the entire sale inVenezuela. In other countries it is being phasedout; in Mexico City lead concentrations decreasedby 80% in the period 1990–4.

Most of the air pollution occurs in major urbancenters. In 1994, 43 of them had more than one mil-lion inhabitants. Often the situation is aggravatedby the cities (e.g. Mexico City and Santiago deChile) being situated in valleys surrounded bymountains.

Not surprisingly, the most critical air pollu-tants in Mexico City are ozone and its precursorsNO2, VOCs, and PM. The ambient ozone concen-trations have consistently exceeded the Mexicanone-hour standard of 220mgm-3. The highest valueever recorded (in 1992) was 955mgm-3.

In Santiago the most critical air pollutant is par-ticulates, especially in the colder period(April–September), the principal source being alarge number of poorly maintained diesel buses(GEMS/AIR 1996). The concentration of TSP isamong the highest in any urban area in the world. In 1995 it reached a one-hour mean of 621mgm-3, compared with the Chilean standard of260mgm-3.

São Paulo in Brazil is the third of the three mostpolluted cities in the region. There has been somesuccess in attempts to control emissions from therapidly growing industry, and SO2 concentrationswere reduced substantially during the 1980s(WHO/UNEP 1992), but ambient air qualitystandards for all traffic relevant pollutants areexceeded.

Africa

For the larger African cities —Cairo, Alexandria,


Nairobi, and Johannesburg —air pollution has beenmonitored for some years and there is an increas-ing awareness of the need for air quality manage-ment (WHO/UNEP 1992; GEMS/AIR 1996).

Urbanization is, however, increasing rapidly allover Africa and especially in the least developedcountries, by up to 5% per year, the driving forcebeing a mixture of population growth, natural dis-asters, and armed ethnic conflicts. Most Africancities have been unable to keep pace with thisdevelopment and lack adequate industrial andvehicle pollution control.

Much of the urban population growth is incoastal cities, e.g. in the Mediterranean area. So fargeneral air pollution appears to be modest, buturban problems are emerging. In most countries,however, emission inventories are nearly nonexistent, pollution is neither monitored norcontrolled, and there are no long-term records ofpollution levels and impacts (UNEP 1997).

West Asia

In the past two decades West Asia has been radical-ly transformed, with an urban growth rate above4% per year. Although the most pressing urbanproblem seems to be waste management, air pollu-tion is also emerging (UNEP 1997). In many casesprotective trade regimes and a lack of environmen-tal regulations have prevented adequate replace-ment of outdated polluting industries. Fuels withhigh sulfur content and old inefficient automo-biles using leaded gasoline have exacerbated urbanair pollution. Recently, however, the situation hasbeen improving. In Oman new industries will beregulated with environmental standards.

East Asia

East Asia contains three of the world’s largestcountries (China, India, and Indonesia), severalminor landlocked states, and a series of islandstates (including the highly industrialized Japan).The region is 35% urbanized and contains abouthalf of the largest cities in the world. It thereforerepresents all stages of pollution development.Urbanization is not restricted to the continent

and the major island states, but is also seen as in-migration to the main island on small islandstates, e.g. in the Maldives. Urban congestion andair pollution is seen as a high-priority problem inmany countries, such as China, India, Pakistan, In-donesia, Philippines, and Thailand (UNEP 1997).

Taken as a whole the rapid growth of energy use,combined with extensive use of coal in most ofAsia, has resulted in a drastic increase in emissionsof sulfur dioxide. Attempts to solve local problemsby installing taller stacks only transformed theminto an extensive regional pollution, and the acidi-fication phenomena, known from North Americaand Europe, are now emerging.

In the important agricultural and industrial re-gion of the Chinese Jiangsu province and Shanghaimunicipality SO2 emissions are already high andare projected to double by the year 2010 (Chang etal. 1998). Model calculations demonstrate that inlarge regions the WHO guideline for long-term ex-posure (40–60mgm-3) is exceeded, as is in some re-gions even the one-hour guideline (350mgm-3).Without drastic measures the short-term guide-line will by 2010 be exceeded in a large part of theprovince for more than 5% of the time. In line withthis the new version of the Atmospheric PollutionControl Law passed by the National People’sCongress in 1995 calls for reductions of emissionsfrom power plants and other large coal users basedon a permit system and emission taxes (Chang etal. 1998).

Aerosol analysis in 1987–92 by a privately es-tablished network (Hashimoto et al. 1994) demon-strated TSP levels for Chinese cities up to averagesabove 500mgm-3 (Lanzhou), to be compared with atentative WHO guideline of 120mgm-3 as a one-hour average. Lower values in Seoul (around 70mgm-3) showed a gradual increase.

Some of the highest reported levels of TSP, com-pared to what is seen in European cities, should,however, be evaluated critically, since not all is ofhuman origin; high concentrations observed inBeijing in the December–April period are partlydue to sand storms from the northern desert. Irre-spective of the origin these particles are generallylarger (Section 15.2.1) and thus have minor healthimpacts (Section 15.4.1).

432 jes fenger

A further complication, with problems forabatement strategies, is forest fires, which are gen-erally blamed for causing smog over major cities inthe ASEAN region (Hassan et al. 1997).

Urban transport is an increasing problem,which has been treated in a series of reports from the World Bank (Walsh 1996; Walsh & Shah1997). A special aspect is the extensive use of mo-torbikes (Fig. 15.4), partly because they are thecheapest means of individual motorized transportfor the expanding working class, partly becausemany Asian cities are too crowded to allow adrastic expansion of the automobile fleet. In 1992motorcycles accounted for 27% of the vehicle fleetin Beijing, and for 65% in Guangzhou. This results in comparatively large emissions in relation to fuel consumption, since most motorcycles have two-stroke engines with poor pollution characteristics.

On a rapid path to industrialization, Taiwan ischaracterized by limited land and fast economicgrowth. Consequently, the environmental stress isserious. In the early 1990s the monthly SO2 aver-ages in Taipei were reported to be about 30 p.p.b.(80mgm-3), but a gradual decrease is expected con-current with a reduction in permissible sulfur con-tents in fuels. The limit for diesel fuel was reducedto 0.05% in 1999. Lead in gasoline is also beingphased out (Fang & Chen 1996).

In a series of major cities, where energy pro-duction is based on gas or low-sulfur coal (e.g.Bombay, Calcutta, Bangkok), SO2 is not a seriousproblem, with average levels about 30mgm-3. Inmany cases, however, TSP levels are above WHOguidelines.

15.7 CONCLUSIONS

Urban air pollution and its impact on urban airquality are a worldwide problem. This manifestsitself differently in different regions, dependingupon economic, political, and technological devel-opment, upon the climate and topography, andlast —but not least —upon the nature and quality ofthe available energy sources. Nevertheless, a se-ries of general characteristics emerges.

15.7.1 From space heating to traffic

Originally urban air pollution was a strictly localproblem mainly connected with space heating andprimitive industries. In the earlier stages of mod-ern industrialization it was considered unavoid-able or even a symbol of growth and prosperity.The situation in the industrialized western worldhas in most respects proved this viewpoint outdat-ed. Emissions from industry and space heating areby and large controllable, but the urban atmos-phere is now in most cities dominated by trafficemissions, with documented impacts on humanhealth. Attention has thereby been shifted fromsulfur dioxide and soot to nitrogen oxides, thewhole spectrum of organic compounds, and particles of various sizes and composition, whichare reported to be carcinogenic and/or cause a sig-nificant reduction in life expectancy through res-piratory and cardiovascular diseases. Thesepollutants require much more detailed investiga-tions in the form of chemical analysis and comput-er modeling.

The nature of material damage has shifted. Inmany respects the problems are diminishing, withreduced levels of sulfur dioxide, the introductionof more resistant materials, and the filtration of in-door air. On the other hand, the interaction be-tween different pollutants and their impacts on,for example, sensitive electronic equipment arenot fully understood.

In principle, control of emissions of sulfur andnitrogen oxides is relatively straightforward whenthey are related to power production in largeplants, which can be compelled to use clean fuelsand equipped with proper cleaning technology.Traffic emissions are more difficult to control,since they, in the nature of things, arise from smallunits. Meeting the increasingly stringent air pollu-tion targets is therefore not an easy task. Accordingto the conclusion of the “Auto Oil Programme,”even with the maximum technical package intro-duced in the EU, not all cities will be able tocomply.

The situation in developing countries is mixed.In some major cities in Asia sulfur emissions havebeen brought under control, e.g. via a transition


from coal to natural gas, but particularly in China arapid growth in energy production based on coalhas resulted in increasing sulfur pollution on bothurban and regional scales. In the developing coun-tries, however, and in some economies in transi-tion (including eastern Europe) traffic is becomingthe problem. This is a challenge to city planning inthese countries, where the long repressed wishesfor private automobiles are difficult to reconcilewith environmental protection.

15.7.2 Regional impacts on urban air pollution

The interactions between the cities and their sur-roundings are becoming increasingly important.With expanding and often merging urban areas anddiminishing emissions in the cities proper, pollu-tion levels can to a large extent be determined bylong-range transport. The same applies to lead pol-lution in countries where lead has been removedfrom gasoline. Another example is photochemical air pollution, which in many cases is a large-scalephenomenon, where emissions and atmosphericchemical reactions in one country may influenceurban air quality in another.

15.7.3 Cities as sources of pollution

A more far-reaching problem is the city as a sourceof pollution. In the past, local problems were at-tacked by dispersing pollutants from high stacks,but this only resulted in a transfer to a larger geo-graphic scale in the form of acidification and othertransboundary phenomena. Now long-lived green-house gases, and especially carbon dioxide, threat-en the global climate —irrespective of their origin.This problem can only be solved by a generalreduction in net emissions.

In the industrialized countries developments intechnology and legislation to protect air and waterquality have in many cases resulted in improvedenergy efficiency and emission reductions, al-though some means of improving urban air quali-ty, such as catalytic converters, which consumeenergy and emit nitrous oxide (another greenhousegas), are contrary to this goal. On a global basis, thegrowing population and its demand for a higher

material standard of living have so far counteract-ed any reductions in total emissions. Therefore,responsibility for the future is both national andglobal, with many actors, comprising nationalenvironmental agencies, international organiza-tions, the World Bank, etc.

Long-term abatement, however, has been in-tensified in recent years. An example is the ICLEIinitiative, with the purpose of achieving andmonitoring improvements in global environmen-tal conditions through cumulative local actions.

15.7.4 A comprehensive approach

The realization that traffic is rapidly becoming theurban air quality problem in both industrializedand developing countries calls for comprehensivesolutions, where traffic-related air pollution isseen in connection with other impacts of traffic,such as noise, accidents, congestion, and generalmental stress. As a consequence, technologicalimprovements in the form of less polluting vehi-cles are not sufficient. Support for infrastructure,where the need for transport is minimized, andwhere use of public means of transport dominatesuse of private automobiles and motorbikes, shouldbe encouraged. Unfortunately, most attempts atcontrol will be perceived as a restriction of individ-ual freedom, and they are frequently met with out-spoken opposition. Obviously, a change in attitudeis called for.

ACKNOWLEDGMENTS

The author thanks colleagues who have providedmaterial and offered advice during the preparationof this chapter —and especially Ole Hertel andFinn Palmgren. Further, he thanks the librarians at the National Environmental Institute, BirgitLarsen and Kit Andersen, and the lithographicartist Britta Munter.

REFERENCES

The literature relevant to urban-scale air pollution is

434 jes fenger

vast, and the references below are only given as typicalexamples. Some of the presented data have been used bydifferent authors and are compiled from various sources.Not all references are therefore primary, but they are keyto a full documentation. Note that many institutionspublish series of reports. New Issues may have appearedafter finalization of this chapter.

Adamson, S., Bates, R., Laslett, R. & Pototschnig, A.(1996) Energy Use, Air Pollution, and EnvironmentalPolicy in Krakow. Can Economic Incentives ReallyHelp? World Bank Technical Paper no. 308. The WorldBank, Washington DC.

Andersen, D. (1988) The Everyday Life of the DanishPopulation 1987 (in Danish). Institute of Social Sci-ence, Copenhagen.

Anderson, H.R., Spix, C., Medina, S. et al. (1997) Airpollution and daily admissions for chronic pulmonarydisease in 6 European cities: results from the APHEA project. European Respiration Journal 10,1064–71.

Anfossi, D. & Sandroni, S. (1997) Ozone levels in Parisone century ago. Atmospheric Environment 31,3481–2.

Arroyave, C. & Morcillo, M. (1995) The effect of nitrogenoxides in atmospheric corrosion of metals. CorrosionScience 37, 293–305.

Ashmore, M. (1998) Impacts on urban vegetation andecosystems. In Fenger, J. et al. (eds), Urban Air Pollu-tion, European Aspects. Kluwer Academic Publishers,Dordrecht.

Bascom, R., Bromberg, P.A., Daniel, A. et al. (1996)Health effects of outdoor air pollution. American Jour-nal of Respiratory Critical Care Medicine 153, 3–50,477–98.

Bellander, T., Svartengren, M., Berglind, N., Staxler, L. &Järup, L. (1999) SHAPE. The Stockholm Study onHealth Effects of Air Pollution and Their EconomicConsequences. Vägverket, Borlänge.

Berkowicz, R. (1998) Street scale models. In Fenger, J. etal. (eds), Urban Air Pollution, European Aspects.Kluwer Academic Publishers, Dordrecht.

Berkowicz, R. (2000) A simple model for urban back-ground pollution. Environmental Monitoring andAssessment.

Bitan, A. (1992) The high climatic quality city of thefuture. Atmospheric Environment 26B, 313–29.

Bollen, J.C., Hettelingh, J.-P. & Maas, R.J.M. (1996) Sce-narios for Economy and Environment in Central andEastern Europe. RIVM report no.481505002. Nether-lands Institute of Public Health and the Environment.

Brimblecombe, P. (1977) London air pollution. Atmos-pheric Environment 11, 1157–62.

Brimblecombe, P. (1987) The Big Smoke. A History of AirPollution in London since Medieval Times. Methuen,London.

Brimblecombe, P. (1990) The composition of museumatmospheres. Atmospheric Environment 24B, 1–8.

Brown, K.S. (1997) A decent proposal? EPA’s new clean air standards. Environmental Health Perspectives 105,378–83.

Brown, L.R., Renner, M. & Flavin, C. (1998) Vital Signs1998. The Environmental Trends that Are ShapingOur Future. W.W. Norton & Company, Inc. New York.

Brown, V.J. (1999) The worst of both worlds: poverty andpolitics in the Balkans. Environmental Health Per-spectives 107, A606–13.

Brunekreef, B. (1997) Air pollution and life expectancy: isthere a relation? Occupational and EnvironmentalMedicine 54, 781–4.

Brunekreef, B. (1999) All but quiet on the particulatefront. American Journal of Respiratory Critical CareMedicine 159, 354–6.

Chang, Y.S., Arndt, R.L., Calori, G., Carmichael, G.R.,Streets, D.G. & Sue, H.P. (1998) Air quality impacts asa result of changes in energy use in China’s Jiangsuprovince. Atmospheric Environment 32, 1383–95.

Cocheo, V., Sacco, P., Boaretto, C. et al. (2000) Urban ben-zene and population exposure. Nature 40, 141.

CORINAIR (1996) Atmospheric Emission InventoryGuidebook. European Environment Agency, Copen-hagen.

Davidson, C.I. (1979) Air pollution in Pittsburgh: A his-torical perspective. Air Pollution Control Association29, 1035–41.

Dennis, R.L., Byon, W.D., Novak, J.H., Galuppi, K.J. &Coats, C.J. (1996) The next generation of integrated airquality modeling: EPA’s models-3. Atmospheric Envi-ronment 30, 1925–38.

Dockery, D.W., Pope, C.A. III, Xu, X. et al. (1993) Anassociation betweeen air pollution and mortality in six US cities. New England Journal of Medicine 239,1753–9.

Edwards, L. (1998) Limit values. In Fenger, J. et al. (eds),Urban Air Pollution, European Aspects. Kluwer Acad-emic Publishers, Dordrecht.

EEA (1998) Europe’s Environment. The Second Assess-ment. European Environment Agency, Copenhagen.

Eggleston, S., Hackman, M.P., Heyes, C.A., Irwin, J.G.,Timmis, R.J. & Williams, M.L. (1992) Trends in urbanair pollution in the United Kingdom during recentdecades. Atmospheric Environment 26B, 227–39.


Ellis, J. & Tréanton, K. (1998) Recent trends in energy-related CO2 emissions. Energy Policy 26, 159–66.

EMEP/MSC-W (1998) Transboundary Acidifying AirPollution in Europe. Research Report no. 66. Norwe-gian Meteorological Institute, Oslo.

EMEP/MSC-W (1999) Transboundary Acidifying AirPollution in Europe. Research Report no. 83. Norwe-gian Meteorological Institute, Oslo.

ENS (1997) Energy Statistics (in Danish). Danish EnergyAgency, Copenhagen.

Enya, T., Suzuki, H., Watanabe, T., Hirayama, T. &Hisamatsu, Y. (1997) 3-Nitrobenzanthone, a powerfulbacterial mutagen and suspected human carcinogenfound in diesel exhaust and airborne particles. Envi-ronmental Science and Technology 31, 2772–6.

European Commission (1995) Towards Fair and EfficientPricing in Transport. COM(95)691. European Commu-nities, Luxembourg.

European Commission (1996a) The European Auto OilProgramme. A report by the Directorate Generals forIndustry, Energy, and Environment. European Com-munities, Luxembourg.

European Commission (1996b) Air Quality Report of theEuropean Auto Oil Programme. Report of Sub Group2. European Communities, Luxembourg.

European Commission (1998) Economic Evaluation ofAir Quality Targets for Sulfur Dioxide, Nitrogen Diox-ide, Fine and Suspended Particulate Matter and Lead.European Communities, Luxembourg.

Faiz, A., Weaver, C.S. & Walsh, M.P. (1996) Air Pollutionfrom Motor Vehicles. Standards and Technologies forControlling Emissions. World Bank, Washington, DC.

Fang, S.-H. & Chen, H.-W. (1996) Air quality and pollu-tion control in Taiwan. Atmospheric Environment 30,735–41.

Fenger, J., Hertel, O. & Palmgren, F. (eds) (1998) UrbanAir Pollution. European Aspects. Kluwer AcademicPublishers, Dordrecht.

Finlayson-Pitts, B. & Pitts, J.N. Jr (1997) Tropospheric air pollution, ozone, airborne toxics, polycyclicaromatic hydrocarbons, and particles. Science 276,1045–52.

Friedrich, R. & Schwarz, U.-B. (1998) Emission inven-tories. In Fenger, J. et al. (eds), Urban Air Pollution,European Aspects. Kluwer Academic Publishers,Dordrecht.

Gaardmand, A. (1993) Land Use Planning in Denmark1938–1992 (in Danish). Arkitektens Forlag, Copen-hagen.

GEMS/AIR (1996) Air Quality Management and Assess-ment in 20 Major Cities. United Nations Environment

Programme and World Health Organization, NewYork.

Goklany, I.M. (1995) Richer is cleaner. In Bailey, R. (ed.),The True State of the Planet. The Free Press, New York.

Goldreich, Y. (1995) Urban climate studies in Israel —areview. Atmospheric Environment 29, 467–78.

Grosjean, D., Grosjean, E. & Williams, E.L. II (1993) Fad-ing of artists’ colorants by a mixture of photochemicaloxidants. Atmospheric Environment 27A, 765–72.

Grossman, G.M. & Krueger, A.B. (1995) Economicgrowth and the environment. The Quarterly Journal ofEconomics 110, 353-377.

Guicherit, R. & van Dop, H. (1977) Photochemical pro-duction of ozone in Western Europe (1971–1975) andits relation to meteorology. Atmospheric Environ-ment 11, 145–55.

Hasford, B. & Fruhmann, G. (1998) Air pollution anddaily admissions for chronic obstructive pulmonarydisease in six European cities: results from the APHEAproject. European Respiration Journal 11, 992–3.

Hashimoto, Y., Sekine, Y., Kim, H.K., Chen, Z.L. & Yang,Z.M. (1994) Atmospheric fingerprints of East Asia,1986–1991. An urgent record of aerosol analysis by the Jack Network. Atmospheric Environment 28,1437–45.

Hassan, H.A., Taha, D., Dahalan, M.P. & Mahmud, A.(eds) (1997) Transboundary Pollution and the Sustain-ability of Tropical Forests. Asean Institute of ForestManagement, Kuala Lumpur.

HLU (1994) Air Quality in the Copenhagen Area (inDanish). Environmental Protection Agency,Copenhagen.

Holgate, S.T., Samet, J.M., Koren, H.S. & Maynard, R.L.(eds) (1999) Air Pollution and Health. Academic Press,London.

Horvath, H. (1998) Reduction of visibility. In Fenger, J. etal. (eds), Urban Air Pollution, European Aspects.Kluwer Academic Publishers, Dordrecht.

ICLEI (2000) Recent Publications. The InternationalCouncil For Local Environmental Initiatives(http://www.iclei.org/iclei/icleipub.htm).

Jantunen, M.J., Hänniken, O., Katsouyanni, K. et al.(1998) Air pollution exposure in European cities. The“EXPOLIS” study. Journal of Exposure Analysis andEnvironmental Epidemiology 8, 495–518.

Jensen, F.P. & Fenger, J. (1994) The air quality in Danishurban areas. Environmental Health Perspectives102(suppl. 4), 55–60.

Jensen, S.S. (1998) Mapping human exposure to traffic airpollution using GIS. Journal of Hazardous Materials61, 385–92.

436 jes fenger

Jol, A. & Kielland, G. (eds) (1997) Air Pollution in Europe1997. EEA Environmental Monograph no. 4. EuropeanEnvironment Agency, Copenhagen.

Kallos, G. (1998) Regional/mesoscale models. In Fenger,J. et al. (eds), Urban Air Pollution, European Aspects.Kluwer Academic Publishers, Dordrecht.

Kassomenos, P., Skouloudis, A.N., Lykoudis, S. & Flocas,H.A. (1999) “Air-quality indicators” for uniform in-dexing of atmospheric pollution over large metro-politan areas. Atmospheric Environment 33, 1861–79.

Katsouynni, K., Touloumi, G., Spix, C. et al. (1997) Shortterm effects of ambient sulfur dioxide and particulatematter on mortality in 12 European cities: results fromtime series data from the APHEA project. British Med-ical Journal 314, 1658–63.

Kemp, K. & Palmgren, F. (1999) The Danish air qualitymonitoring programme. Annual Report for 1998.NERI Technical Report no. 296. National Environ-mental Research Institute, Roskilde.

Kemp, K., Palmgren, F. & Mancher, O.H. (1998) The Dan-ish air quality monitoring programme. Annual Reportfor 1997. NERI Technical Report no. 245. NationalEnvironmental Research Institute, Roskilde.

Komeiji, T., Aoki, K. & Koyama, I. (1990) Trends of airquality and atmospheric deposition in Tokyo. Atmos-pheric Environment 24A, 2099–103.

Krawack, S. (ed.) (1991) Air pollution from individual andpublic transport (in Danish). Miljøprojekt 165. DanishEPA, Copenhagen.

Krupnick, A.J. & Portney, P.R. (1991) Controlling urbanair pollution: a benefit–cost assessment. Science 252,522–8.

Kucera, V. & Fitz, S. (1995) Direct and indirect air pollution effects on materials including cultural mon-uments. Water, Air and Soil Pollution 85, 153–65.

Kucera, V., Pearce, D. & Brodin, Y.-W. (eds) (1996) Eco-nomic Evaluation of Air Pollution Damage to Materi-als. Swedish EPA, Stockholm.

Kurashige, Y. & Miyashita, A. (1998) How many days canMt.Fuji and the Tokyo Tower be seen from the Tokyosuburban area. Journal of the Air and Waste Manage-ment Association 48, 763–5.

Larsen, J.C. & Larsen, P.B. (1998) Chemical carcinogens.In Hester, R.E. & Harrison, R.M. (eds), Air Pollutionand Health. Issues in Environmental Science andTechnology 10. The Royal Society of Chemistry,Cambridge.

Larssen, S. & Hagen, L.O. (1997) Air Pollution Monitor-ing in Europe —Problems and Trends. Topic report 26.European Environment Agency, Copenhagen.

Lynn, W.R. (1999) Megacities. Sweet dreams or environ-mental nightmares. Environmental Science and Tech-nology News 238A–40A.

Mage, D., Ozolins, G., Peterson, P. et al. (1996) Urban airpollution in megacities of the world. Atmospheric En-vironment 30, 681–6.

Mage, D., Wilson, W., Hasselblad, V. & Grant, L. (1999)Assessment of human exposure to ambient particulatematter. Journal of the Air and Waste Management As-sociation 49, 1280–91.

Mamane, Y. (1987) Air pollution control in Israel duringthe first and second century. Atmospheric Environ-ment 21, 1861–3.

Marico, M.M., Podsiadlik, D.H. & Chase, R.E. (1999) Ex-amination of size-resolved and transient nature ofmotor vehicle particle emissions. EnvironmentalScience and Technology 33, 1618–26.

Mayerhofer, P., Weltschev, M., Trukenmüller &Friedrich, R. (1995) A methodology for the economicassessment of material damage caused by SO2 and NOxemissions in Europe. Water, Air and Soil Pollution 85,2687–92.

Maynard, R.L. & Howard, C.V. (eds) (1999) ParticulateMatter. Properties and Effects upon Health. Bios Sci-entific Publishers, Oxford.

McMichael, A.J., Anderson, H.R., Brunekreef, B. &Cohen, A.J. (1998) Inappropriate use of daily mortalityanalyzes to estimate longer-term mortality effects ofair pollution. International Epidemiological Associa-tion 27, 450–3.

Mestayer, P.G. & Anquetin, S. (1995) Climatology ofcities. In Gyro, A. & Rys, F.-S. (eds) Diffusionand Transport of Pollutants in AtmosphericMesoscale Flow Fields. Kluwer Academic Publishers,Dordrecht.

Newby, P.T., Mansfield, T.A. & Hamilton, R.S. (1991).Sources and economic implications of building soilingin urban areas. Science of the Total Environment 100,347–65.

Nielsen, H.P. (1997) Localization, transport means andurban structure (in Danish). Byplan 6/97.

Nielsen, T. (1996) Traffic contribution of polycyclic aro-matic hydrocarbons in the center of a large city. At-mospheric Environment 20, 3481–90.

Nielsen, T., Feilberg, A. & Binderup, M.L. (1999) Thevariation of street air levels of PAH and other muta-genic PAC in relation to regulations of traffic emis-sions and the impact of atmospheric processes.Environmental Science and Pollution Research 6,133–7.


Novak, D.J., Civerolo, K.L., Rao, S.T., Sistla, G., Luley,C.J. & Crane, D.E. (2000) A modeling study of the im-pact of urban trees on ozone. Atmospheric Environ-ment 34, 1601–13.

Nriagu, J.O. (1990) The rise and fall of leaded gasoline.Science of the Total Environment 92, 13–28.

OECD (1995a) Motor Vehicle Pollution. ReductionStrategies beyond 2010. Organization for EconomicCo-operation and Development, Paris.

OECD (1995b). Urban Travel and Sustainable Develop-ment. Organization for Economic Co-operation andDevelopment, Paris.

OECD (1998) Advanced Air Quality Indicators and Re-porting. Organization for Economic Co-operation andDevelopment, Paris.

OECD (1999) OECD Environmental Data. Compendi-um 1999. Organization for Economic Co-operationand Development, Paris.

O’Meara, M. (1999) Reinventing Cities for People and thePlanet. Worldwatch Paper 147.

Onursal, B. & Gautam, S.P. (1997) Vehicular Air Pollu-tion. Experiences from Seven Latin American UrbanCenters. World Bank Technical Paper no. 373. WorldBank, Washington, DC.

Palmgren, F., Berkowicz, R., Hertel, O. & Vignati, E.(1996) Effects of reduction of NOx on NO2 levels inurban streets. Science of the Total Environment189/190, 409–15.

Phillips, D.R. (1993) Urbanization and human health.Parasitology 106, S93–107.

Pichl, P. (ed.) (1998) Exchange of Experiences betweenEuropean Cities in the Field of Clean Air Planning,CO2 Reduction and Energy Concepts. Umwelt BundesAmt, Berlin.

Pope, C.A. III, Dockery, D.W. & Schwartz, J. (1995) Re-view of epidemiological evidence of health effects ofparticulate air pollution. Inhalation Toxicology 7,1–18.

Population Reference Bureau (1999) World PopulationData Sheet. Population Reference Bureau, Washing-ton, DC.

Renner, R. (2000) Pollution monitoring should get per-sonal, scientists say. Environmental Science and Tech-nology News 64–65A.

Richter, D.U.R. & Williams, W.P. (1998) Assessment andManagement of Urban Air Quality in Europe. EEAMonograph no. 5. European Environment Agency,Copenhagen.

Schmidt, A.L., Bronée, P., Kemp, K. & Fenger, J. (1999)Airborne dust on museum exhibits. Paper presented to

Air Quality in Europe: Challenges for the 2000s,Venice, May 19–21.

Schwartz, J., Dockery, D.W. & Neas, L.M. (1996) Is dailymortality associated specifically with fine particles?Journal of the Air and Waste Management Association46, 927–39.

Schwela, D. & Zali, O. (eds) (1999) Urban Traffic Pollu-tion. E & FN Spon, London.

Selden, T.M., & Song, D. (1994) Environmental qualityand development: is there a Kuznets curve for air pollu-tion emissions? Journal of Environmental Economicsand Management 27, 147–62.

Shafik, S. (1994) Economic development and environ-mental quality: an econometric analysis. Oxford Eco-nomic Papers 46, 757–73.

Skov, H., Christensen, C.S., Fenger, J., Larsen, D. &Sørensen, L. (2000) Exposure to indoor air pollution ina reconstructed house from the Danish Iron Age.Atmospheric Environment 34, 3801–4.

Smith, K.R. (1988) Air pollution. Assessing total expo-sure in developing countries. Environment 30, 16–20,28–30, 33–5.

Smith, K.R. (1993) Fuel combustion, air pollution expo-sure, and health: the situation in the developing coun-tries. Annual Review of Energy and the Environment18, 529–66.

Spix, C., Anderson, H.R., Schwartz J. et al. (1998) Short-term effects of air pollution on hospital admission ofrespiratory diseases in Europe: a quantitative summa-ry of APHEA study results. Archives of EnvironmentalHealth 53, 54–64.

Stanners, D. & Bourdeau, P. (eds) (1995) Europe’s Envi-ronment: The Dobris Assessment. European Environ-ment Agency. Office for Publications of the EuropeanCommunities, Luxemburg.

Thomas, V.M., Socolow, R.H., Fanelli, J.T. & Spiro, T.G.(1999) Effects of reducing lead in gasoline: an interna-tional experience. Environmental Science and Tech-nology 33, 3942–8.

Tidblad, J. & Kucera, V. (1998) Materials damage. InFenger, J. et al. (eds), Urban Air Pollution, EuropeanAspects. Kluwer Academic Publishers, Dordrecht.

Tuch, Th., Heyder, J., Heinrich, J. & Wichmann, H.E.(1997) Changes of the Particle Size Distribution in anEastern German City. Presented at the Sixth Interna-tional Inhalation Symposium, Hannover MedicalSchool.

UN (1997a) Report of the United Nations Conference onHuman Settlements (Habitat II). United Nations,New York.

438 jes fenger

UN (1997b) Statistical Yearbook 1995. United Nations,New York.

UNEP (1991) Urban Air Pollution. UNEP/GEMS Envi-ronment Library No 4. United Nations EnvironmentProgramme, Nairobi.

UNEP (1997) Global Environment Outlook. UnitedNations Environment Programme, Nairobi.

UNEP/WHO (1994) GEMS/AIR Methodology ReviewsVolume 3: Measurement of Suspended ParticulateMatter in Ambient Air. United Nations EnvironmentProgramme, Nairobi.

Vignati, E., Berkowicz, R. & Hertel, O. (1996) Compari-son of air quality in streets of Copenhagen and Milan,in view of the climatological conditions. Science of theTotal Environment 189/190, 467–73.

Volz, A. & Kley, D. (1988) Evaluation of the Montsourisseries of ozone measurements made in the nineteenthcentury. Nature 332, 240–2.

Wallace, L. (1996) Indoor particles: a review. Journal of the Air and Waste Management Association 46,98–126.

Walsh, M.P. (1996) Motor vehicle pollution control inChina: an urban challenge. In Stares, S. & Zhi, L. (eds),China’s Urban Development Strategy. World BankDiscussion Paper no. 353. World Bank, Washington,DC.

Walsh, M. & Shah, J.J. (1997) Clean Fuels for Asia. Tech-nical Options for Moving toward Unleaded Gasolineand Low-sulfur Diesel. World Bank Technical Paperno. 377. World Bank, Washington, DC.

Whitby, K.T. & Sverdrup, G.M. (1980) California

aerosols: their physical and chemical characteristics.Advances in Environmental Science and Technology9, 477–517.

WHO/UNEP (1992) Urban Air Pollution in Megacities ofthe World. Blackwell, Oxford.

WHO (1995) Concern for Europe’s Tomorrow: Healthand Environment in the WHO European Region.Wissenschaftliche Verlag, Stuttgart.

WHO (1987/2000) Air Quality Guidelines for Europe.World Health Organization, Copenhagen.

Wiederkehr, P. & Yoon, S.-J. (1998) Air quality indicators.In Fenger, J. et al. (eds), Urban Air Pollution, EuropeanAspects. Kluwer Academic Publishers, Dordrecht.

Wilson, W.E. & Suh, H.H. (1997) Fine particles and coarseparticles: concentration relationships relevant toepidemiologic studies. Journal of the Air and WasteManagement Association 47, 1238–49.

World Commission on Environment and Development(1987) Our Common Future. Oxford University Press,Oxford.

Zeger, S.L., Dominici, F. & Samet, J. (1999) Harvesting-resistant estimates of air pollution effects on mortali-ty. Epidemiology 10, 171–5.

Zhiqiang, Q., Siegmann, K., Keller, A., Matter, U., Sher-rer, L. & Siegmann, H.C. (2000) Nanoparticle air pollu-tion in major cities and its origin. AtmosphericEnvironment 34, 443–51.

Zlatev, Z., Fenger, J. & Mortensen, L. (1996) Relation-ships between emission sources and excess ozone con-centrations. Computers Mathematics Application 32,101–23.

handbook of atmospheric science || urban-scale air pollution

Documents