Handbook of Atmospheric Science || Regional-Scale Pollution Problems

14.1 INTRODUCTION

It is only in the past 40 years that evidence has arisen to show that air pollution poses a problem inareas well removed from obvious sources such ascities and associated industry. Pollutants in ambi-ent urban air will be rapidly dispersed and diluteddownwind of the conurbation. This fact, followingthe catastrophic smog events of the early 1950s in London, led industry —particularly the powerindustry —to build taller stacks to aid the dispersalof “smoke” away from the local environment. Theobservation that pollutants were carried by theprevailing winds and remained airborne for longenough to affect areas very far removed from theimmediate source was only made in the 1960swhen the effects of acid precipitation became apparent on Scandinavian lakes. This was the firsttime that transboundary air pollution was recog-nized as a significant hazard, which prompted legislation in the form of the United Nations Economic Commission on Europe (UNECE) to instigate the Convention on Long-range Trans-boundary Air Pollution (CLRTAP). The con-vention was the first internationally legally binding instrument to deal with problems of airpollution on a broad regional basis. It was signed in 1979 and entered into force in 1983. The firstprotocol under this convention focused on sulfurdioxide (SO2), and in the past 20 years a number of protocols have been ratified for a variety of pollutants. (Further details of specific pro-tocols can be found at the UNECE website

http://www.unece.org/env/lrtap/). Table 14.1 pro-vides a summary of these protocols along with the year that they were ratified. It is the imple-mentation of these protocols (at the national level) that has driven regional-scale air pollutionmanagement.

The term “regional air pollution” (as opposed tourban or global effects) is a phrase coined largelythrough work on acidic species, as well as on pho-tochemical oxidants such as ozone (O3). The term“regional” implies the widespread dispersal of primary and secondary air pollutants, the latterformed from the oxidation or reaction of primarypollutants released from a variety of local or pointsources. Perhaps the best examples include the oxidation of SO2 to sulfate (SO4

2-), or the oxidationof nitrogen dioxide (NO2) to nitrate (NO3

-). Theformation of nitrate and sulfate particulates is a relatively slow process, which occurs during prolonged atmospheric transport, resulting in deposition of acidic species well away from thepoint of release. The occurrence of particulate material released either directly to the atmos-phere or formed through reaction contributes to re-duced visibility or haze, and regional air pollutionis often distinguished by an easy reference such asvisibility. Perhaps the best example is “Arctichaze,” a regional air pollution phenomenonbrought about by long-range transport of air pollu-tants, that was first observed by pilots in the 1950sand is the focus of further attention later in thischapter.

The dimensions to regional-scale air pollution

14 Regional-Scale Pollution Problems

CRISPIN J. HALSALL

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

Copyright © 2003 by Blackwell Publishing Ltd

Regional-Scale Pollution Problems 377

are often difficult to define. However, it is gener-ally accepted that on a spatial scale the area affec-ted ranges from the size of a country up to the subcontinental level. Often “regional scale” isused in a context to define a certain populated orindustrial area where air pollution problems arewell characterized, such as the Los Angeles orAthens basins. However, under scientific criteriaregional air pollution remains at the synoptic scale(thousands of kilometers). The vertical extent ofthe pollution is usually confined to the atmos-pheric boundary layer (~500–2000m), with incur-sions into the free troposphere depending on the prevailing meteorology. Regional pollution pro-blems are generally considered over the period ofdays to months, i.e. on a seasonal time frame. Inci-dents such as ground-level ozone during the sum-mer or wintertime haze in the Arctic are goodexamples of this, but this is clearly dependent onthe nature and source of pollution. Criteria defining temporal and spatial scales of air pollu-tion are presented in Table 14.2. These definitionsare from Borrell et al. (1997) as part of EURO-TRAC, a European framework organization set upin 1988 with the aim of coordinating research onchemical pollutants in the troposphere. The re-gional air pollution criteria were developed withphotooxidants in mind but are applicable to othertypes of pollutant.

14.2 MONITORING FRAMEWORKS

The main framework of the UN-ECE CLRTAP was the setting up and financing of the EMEP program (Cooperative Programme for Monitoringand Evaluation of the Long-range Transport of Air Pollution in Europe). The EMEP program relies onthree main areas to fulfill its role, namely collec-tion of emission data, measurements of air and precipitation quality, and modeling of atmos-pheric transport and deposition of air pollutants.Through the combination of these three elements,EMEP regularly reports on emissions, concentra-tions, and/or deposition of air pollutants, as well as the quantity and significance of transboundaryfluxes and related exceedances to critical loads and

Table 14.1 UNECE Convention on Long-range Transboundary Air Pollution.

Protocol Year Summary

Reduction in sulfur 1985 30% reduction in S emissions by 1993 (base year 1980)

Control of NOx* 1988 Reduction in NOx emissions by 9% (base year 1987)

Control of VOC† 1991 30% reduction in VOC by 1999 (base year between 1984 and 1990)

Further reduction in sulfur 1994 Introduction and setting of critical loads

Control on heavy metals 1998 Reduce emissions for Cd, Pb, and Hg to below 1990 levels

Persistent organic pollutants 1998 Eliminate discharges and emissions of 11 pesticides, two industrial compounds and three

by-products

Abatement of acidification, 1999 Emission rulings for 2010 for NOx, VOC and NH3, i.e. cut

eutrophication and ozone NOx emissions by 41% compared to 1990

*Nitric oxide (NO) and nitrogen dioxide (NO2).

†Volatile organic compounds.

Table 14.2 Scales of air pollution (adapted from Borrellet al. 1997).

Vertical scale Horizontal Time scale (km) scale (km) (days)

Local or urban 2 100 1

Regional or 2 2,500 5–100

continental

Global 10 20,000 100

to years

378 crispin j . halsall

threshold levels. The modeling program has gonesome way to help shape emission reduction targetsfor some of the later protocols. EMEP now operatessome 100 sampling stations across 32 Europeancountries. In the USA, air pollution on both localand regional scales falls under the remit of the Environmental Protection Agency, largelythrough the 1990 Clean Air Act. Air pollutionmonitoring, assessment of source emissions, andair quality criteria for key pollutants now all fallunder EPA control. With regard to regional air pollution the Ambient Monitoring Technology Information Center (AMTIC), operated by theEPA’s Monitoring and Quality Assurance Group(MQAG), contains information and files on ambi-ent air quality monitoring programs, as well as de-tails on monitoring methods and information onair quality trends. Both the US EPA and EMEPmaintain sophisticated websites containing rele-vant monitoring data, as well as model results andpredictions for regional air quality and depositionestimates.

14.3 THE REGIONAL OZONEPROBLEM

Photochemical oxidants, of which O3 is the mostimportant, are secondary air pollutants, formed bythe reaction of volatile organic compounds (VOCs)and carbon monoxide (CO) with nitrogen oxides(NOx) in the presence of sunlight. High levels of O3in the atmospheric boundary layer (ground-levelO3), along with other photooxidants such as hy-drogen peroxide (H2O2) and peroxyacetyl nitrate(PAN), are therefore mainly a summertime phe-nomenon in Europe and North America.

Photochemical smog and associated photo-oxidants were first studied in the Los Angelesbasin of southern California in the late 1940s. Theoccurrence of ground-level O3 at concentrationsgreatly above background levels was established inmost industrialized countries in the followingdecades (e.g. Cox et al. 1975; Isaksen et al. 1978;White et al. 1983; Lui et al. 1987; Dollard et al.1995). With the establishment of continent-widemonitoring networks such as the European EMEP

program and the remit granted to the US EPA, lev-els of ground-level O3 are now routinely monitoredand interpolated, and the extent of the pollution is assessed on a regional basis. Typical Europeanmean rural boundary layer O3 concentrations varywidely, but generally fall between 20 and 40p.p.b.v. (parts per billion by volume) on an annualbasis. Levels during the summer months, how-ever, are higher, between 50 and 60 p.p.b.v., withepisodes often lasting a number of days where concentrations may be well in excess of 80 p.p.b.v.Figure 14.1 shows the average annual maxima inO3 for the period 1988–94 for EMEP, with monitor-ing stations plotted as their latitudinal position.This serves to illustrate the elevated concentra-tions observed in central Europe due to photo-chemical episodes, where concentrations mayreach ~100–150 p.p.b.v., due to stable high-pressure weather systems and an accumulation ofprecursor compounds. These episodes may typi-cally last from 2 to 10 days. The lower values forthe Iberian peninsula reflect the cleaner Atlanticair masses that impact on these sites.

O3is not produced or emitted directly by pollut-ing sources, but formed downwind via a complexreaction of VOC and NOx in the presence of sun-light; a photochemical reaction that is discussed in further detail in the following sections. Conse-quently, O3 can be formed over rural areas. In Europe the highest summertime concentrations ofO3 are found in the southeast of the continent,with lower concentrations to the northwest. Typi-cally, the lowest concentrations occur during thewinter, with a maximum during early summer.

14.3.1 Human health

An important issue surrounding photochemicaloxidants is the human health effects associatedwith acute exposure during photochemical smogevents (PORG 1993). Exposure to high levels of O3can result in respiratory disorders through the irri-tation of the mucous membranes, often accompa-nied by an inflammatory response in the lining ofthe lungs. General sensitization can occur particu-larly for asthmatics or people suffering from otherrespiratory-related allergies. Detailed studies of


the dose–response relationship for O3 have shownthat both the concentration and duration of expo-sure are important factors affecting aspects of lungfunction (Larson et al. 1991, cited in PORG 1997).Of importance is the available evidence on generalpopulation health effects in view of exposure to O3over a wide area. It is suggested that those who livein areas of repeated high O3 concentrations aremore likely to suffer from asthma and have gener-ally depressed lung function (Schwartz 1989). Several studies have shown clear links betweenhospital admissions for asthma/respiratory ill-nesses and levels of O3. However, it is often diffi-cult to distinguish the effects of O3 from those ofother air pollutants (e.g. Oehme et al. 1996). TheWorld Health Organization (WHO) has sum-marized a large number of studies from NorthAmerica and Europe to examine exposure–response relationships. This work calculated a co-efficient for hospital admissions for respiratory disorders, based on O3 alone, to be 0.003810 p.p.b.-1

O3. This allows actual quantification of hospital admissions due to exposure to O3 at the total popu-lation or regional level. Large uncertainties arise,however, once assumptions are used based onwhether a threshold effect or nonthreshold effect is followed, and this can result in the generation of a wide range of admission numbers. Nevertheless,the World Meteorological Organization (WMO) has set guidelines for the protection of humanhealth of between 75 and 100 p.p.b. as maximum

hourly exposure values. Similarly, individual countries have developed their own criteria based on international guidelines. For instance, inthe UK, the Air Quality Guidelines for O3 havebeen set at 50 p.p.b. for an eight-hour running aver-age. In the USA, the National Ambient Air QualityStandard set by the Environment ProtectionAgency for O3 is 120 p.p.b. as a one-hour average, although a secondary standard of 80 p.p.b. for an eight-hour running average has still to be implemented.

14.3.2 Effects on vegetation

Large areas in both Europe and North America areexposed to levels of O3 that exceed a thresholdvalue or critical level whereby potential reduc-tions in crop yield may occur. In addition, O3 mayalso be contributing to forest decline in central Europe (Herman et al. 2001). Acute O3 exposure re-sults in visible damage to the foliar parts of plantswhen rates of O3 uptake (over hours or days) exceedrates of detoxification. Symptoms of acute injuryon broad-leafed plants include bleaching, chloro-sis, and necrosis spots, while in conifers bandingand tip necrosis of the needles are often observed(Krupa et al. 1989). Chronic responses to O3 expo-sure are often more subtle. This results from bothhigh and low rates of O3 uptake throughout thegrowth season or life cycle of the plant. While visible signs of plant injury may occur through

30 35 40 45 50 55

Latitude (degrees)A

vera

ge

annual m

axi

ma, 1

98

8–

94

(p.p

.b.)

60 65 70 75 800

20

40

60

80

100

120

140

SpainPortugal

GermanySloveniaAustria

SwitzerlandItaly

France

FinlandNorwaySweden

UKDenmark

The NetherlandsLithuaniaGermany

NorwayFinlandSweden

Fig. 14.1 Average of 1988–94 annualmaxima at EMEP O3 monitoringsites (plotted according to latitude).(From PORG 1997.)


chronic exposure, reductions in growth or primaryproductivity are the real area of concern.

The benchmark for chronic effects on vegetationis based on the critical levels approach, wherebyconcentrations of O3 above a certain level are likelyto cause an adverse effect on vegetation, such as re-duced crop yield (Grunhage et al. 1999). Establish-ing an appropriate critical level has been the focus of much scientific investigation, with particular attention on crop species. The first critical levelswere derived from American studies, such as theNational Crop Loss Assessment Program, whichestablished the first long-term (or “growing sea-son”) O3 critical level of 25 p.p.b. as a seasonalseven-hour mean (Heck et al. 1988). Concern overthe use of a mean value —which may hide the effectsof high O3 levels —has resulted in the assessment ofdifferent O3 indices. For example, Lee et al. (1988)obtained data from a wide range of studies investi-gating different crop types in a variety of locationsover a number of years. They demonstrated thatthose indices that emphasized peak concentrationsor accumulated exceedances of a threshold concen-tration gave a better fit to crop yield data than meanconcentrations alone. This gave rise to a variety oflong-term experiments to investigate thresholdvalues, perhaps the most extensive being the European open-top chamber programme (EOTCP),which ran over five years (1987–91) in nine coun-tries. Chamber measurements were conducted on avariety of crop types exposed to varying concentra-tions of O3. For the crop species studied, yield reduc-tion was highly correlated with cumulativeexposure above a threshold of either 30 or 40 p.p.b.There was less fit with the data when the 50 p.p.b.threshold was utilized. As background levels of O3are typically ~25–30 p.p.b. in Europe, it was con-sidered that the value of 40 p.p.b. would be the mostappropriate criterion. The exposure index is there-fore referred to as AOT 40, or accumulated exposureover a threshold of 40 p.p.b. The time period overwhich the exposure could be assessed was selectedas three months, effectively the summer months ofMay, June, and July (relevant to northern Europewhen O3 concentrations are at their highest), and also the average length of the EOTCP wheat experiments.

Figure 14.2 shows the relationship between rel-ative grain yield and AOT 40 exposure over threemonths and includes both European and US data.The critical level for agricultural crops has nowbeen selected as 3000 p.p.b. h (above 40 p.p.b. O3),accumulated during daylight hours over the three-month period, which from Fig. 14.2 gives rise to an approximate reduction in wheat yield of 10%.Efforts are now directed at mapping AOT 40 expo-sures across Europe to assess those areas whereambient O3 exposures exceed critical levels; an approach that will influence control strategies forreducing the emissions of O3 precursors. AOT 40 exposure values have also been assessed forother vegetation types; for example, the criticalozone level for forest trees has been defined at anAOT 40 of 10 p.p.b. h during daylight hours over a six-month growing season. However, there is evidence to suggest that some species of tree aremore sensitive to O3 exposure and could suffer significant damage even below this critical level(Van der Heyden et al. 2001).

00

10

20

30

40

50

60

70

80

90

100

110

120

5000

Rel

ati

ve g

rain

yie

ld (

%)

10,000 15,000 20,000

AOT40 (p.p.b. h)

25,000 30,000 35,000

Relative yield (%) = 99.6 – 1.7 ¥ 10–3

AOT40 (p.p.b. h)

R2 = 0.89

99% confidence limits

Fig. 14.2 Relative grain yield of spring wheat and O3 exposure as AOT 40 over three months, based ondata from various European and US open-chamberexperiments (represented by the different symbols).(From Fuhrer 1996.)


Concern over the uncertainties surroundingAOT 40 as an acceptable criterion on which to basea critical level have been highlighted by Fowler et al. (1999). In the EOTCP chamber studies, themeasured reduction of crop yield under controlledconditions was based on crops grown under optimum conditions (i.e. well watered crops)(Manning & Krupa 1992). The crops were also sub-ject to a continuous flow of air with a stable con-centration of O3. This is often in contrast to thesituation in the crop field, where large fluctuationsin O3 may occur over a short time period, affectingboth O3 deposition and plant uptake (Krupa &Kickert 1989). Ongoing studies are focusing on therelationship between measured O3, deposition,and plant uptake at the micrometeorological scale and the subsequent chronic response of theplant.

14.3.3 Understanding the chemistry behindozone formation

The formation of ozone in the polluted atmos-phere is through a complex series of reactions, butin essence relies on the fact that oxidation of nitricoxide (NO) to nitrogen dioxide (NO2) can occurwithout the net consumption of O3, a process thatdoes occur in the unpolluted background atmos-phere. During daylight hours NO2 undergoes pho-tolysis to NO and singlet oxygen (O), the latterreacting with oxygen (O2) to form O3. In turn O3reacts with NO to form NO2. The process leads to the “photostationary steady state,” in which O3formation is matched by O3 loss. This process istypical of the clean background atmosphere wherelevels of NO as well competing oxidants to O3are low. The following reactions illustrate thisprocess:

(14.1)

(14.2)

(14.3)

In the polluted atmosphere, however, the oxida-tion of NO to NO2 can occur via competitive reac-

NO O NO O3 2 2+ Æ +

O O M O2 3+ +( ) Æ

NO NO O2 Æ +

tion with oxidants other than O3, namely the per-oxy radicals of HO2 and RO2. These peroxy radicalsare formed by the reaction of the hydroxyl radical(OH) with VOCs. The oxidation of NO can there-fore be represented by:

(14.4)

(14.5)

Since the conversion of NO to NO2 as a result of reactions 14.4 and 14.5 does not consume O3,the subsequent photolysis of NO2 (reaction 14.1)results in net O3 formation (via reaction 14.2), and the photostationary steady state is disrupted.

It must be pointed out that the above descrip-tion is a much simplified form of a complex seriesof reactions. Importantly, the VOC or hydrocar-bon —usually represented as RH (i.e. alkane) —undergoes a range of oxidation steps resulting inalkoxy radicals (RO), which in turn result in theformation of carbonyls (R-HO). These in turn canbe further oxidized (ultimately to CO2), resultingin further O3 formation with each step. The O3 inturn can be photolyzed, resulting in more OH,causing a chain propagating process. The govern-ing factor in the oxidation of VOCs are the ambientconcentrations of hydroxy (OH), peroxy (HO2), andalkoxy radicals (ROx). To illustrate, Fig. 14.3 pres-ents the schematic of photochemical oxidant pro-duction in the polluted atmosphere (PORG 1997).It must be pointed out, however, that the cycle ofreactions in Fig. 14.3 represents the oxidation of asimple generic alkane to its first generation oxi-dized product R-HO. In fact, the cycle occurs formany different VOCs present in the polluted at-mosphere and is repeated for their oxidized prod-ucts —a process involving many different rates andproducts and hence yields of O3. In an attempt toquantify O3 yield, the photochemical oxidant cre-ation potential (POCP) has been derived for themajor hydrocarbon species present in the atmos-phere, based on their observed concentration andrates of reaction with OH (the dominating oxida-tion process) (Derwent et al. 1996). Elegant de-scriptions of photooxidant and O3 chemistry are

RO NO RO NO2 2+ Æ +

HO NO OH NO2 2+ Æ +


provided in the excellent reviews of Fowler et al.(1999) and Kley et al. (1999).

14.3.4 The importance of VOCs and NOxconcentrations in O3 formation

With regard to regional O3 pollution, attentionmust now be turned toward the O3 precursors ofVOCs and NOx. In the urban atmosphere duringincidents of high sunlight (usually associated withstable anticyclonic conditions), the formation ofO3 is highly sensitive to levels of VOCs and NOxpresent in the atmosphere. For instance, in theurban atmosphere levels of primary pollutantssuch as NO, emitted directly from nearby sourcessuch as vehicle exhausts, are high. These high levels of NO will react not only with HO2 and RO2,but also with O3. This has the effect of actually reducing O3 in the polluted urban atmosphere,

notably for periods of the day when NO levels are high, such as the morning and afternoon rushhours. Levels of O3 may only start to increase oncethe polluted air mass is advected out of a conurba-tion and levels of NO start to decrease. The sensi-tivity of O3 formation is therefore often expressedin terms of the relative ratio of VOC and NOxconcentrations. Situations may arise whereby the VOC/NOx ratio is low, resulting in VOC-limited conditions, whereby a reduction in “RH”will result in a decrease in the reaction chainlength apparent in Fig. 14.3, resulting in a subse-quent decrease in O3. Conversely, a reduction inNOx relative to VOCs may result in NOx-limitedconditions, resulting in plentiful HO2 and RO2radicals that may give rise to further O3 yield.Obviously, further reduction in NOx in a NOx-limited area would ultimately result in O3 decline,but a decline in a highly NOx-polluted environ-

NO2

RO

NO

RO2

O2

HO2

O2

NO2

NO2

O2

HO2

R–HO

HO2HNO3

O3

O3

ROOH

RH

OH

NO

H2O

H2O2

HCHO

Sunlight

Sunlight

Sunlight

Sunlight

Fig. 14.3 Photochemical oxidantproduction in a polluted atmosphere(where NO > 20 p.p.t.v.). Note this isa single cycle for a generic saturatedhydrocarbon “RH.” In anunpolluted atmosphere wherelevels of NO are <20 p.p.t.v., anyRO2 formed from hydrocarbons mayreact with HO2, thus terminatingthe chain reaction that leads to O3 formation. (Reprinted fromEnvironmental Pollution (Fowler etal. 1999). Copyright 1999, withpermission from Elsevier Science.)


ment may give rise to extra O3. The VOC/NOxratio has been estimated to lie in the range of 1–10for the polluted boundary layer over Europe, and insome rural areas due to the emission of biogenicVOCs may increase to 10–20, a situation resultingin NOx-limited conditions and favoring O3 forma-tion (Kley et al. 1999). Figure 14.4 displays a mod-eled two-dimensional cross-sectional profile of O3and NOx concentrations in a transect starting up-wind of an urban area and extending some 150kmdownwind. The formation of O3 is at its highest inthe downwind rural area. This graph effectively il-lustrates the potential for O3 production over awide area, particularly given the multitude ofurban centers across Europe and North America.Similarly, the decline in NOx with increasing dis-tance from the urban source is a result of furtheroxidation, leading to the formation of acidicspecies such as nitric acid (HNO3). Collectivelythese species are termed NOy, and further exam-ples are given in Fig. 14.5.

In order to predict O3 levels, photochemical ox-idant models have been developed in both NorthAmerica and Europe over the past 20 years and vary

according to level of complexity. Essentially, thesedescribe how the primary emitted pollutants aredispersed and chemically transformed as they aretransported away from major source areas. The response in O3 concentrations to changes in pre-cursor emissions (i.e. NOx and VOCs) are often thegoal of these models, and may be used to assess theimpact of regulatory polices. These models oper-ate over varying degrees of spatial resolution (grid-ded scales) and rely on the accuracy of emissioninventories in order to predict ground-level airconcentrations, deposition fluxes and AOT 40 values. As with other air pollutants (see Section14.4.4), both Lagrangian and Eulerian modeling approaches have been utilized for regional photo-chemical oxidant modeling. The Lagrangian models are generally less computationally de-manding, whereby chemical production and re-moval of pollutants occur within a parcel of airadvected over the surface along a trajectory. In theEulerian approach the chemistry and transport aredetermined over a fixed grid for a region of interest;the chemistry and various processes are solved si-multaneously, akin to numerical weather models.

Long-rangetransport

Receptorregion

20 40 80 160

km

km

20 2

NO2, NOy (p.p.b.)

NOy

NO2

O3 (p.p.b.) Peak O3concentration Loss to deposition

100

Sourceregion

Fig. 14.4 Schematic diagramshowing the urban–ruralboundary layer and theformation of O3 and theoxidation of NOx to NOy.(Reprinted from EnvironmentalPollution (Fowler et al. 1999).Copyright 1999, withpermission from ElsevierScience.)


As an example, EMEP operates the Lagrangianozone model, which follows parcels of air along 98-hour long trajectories, with emissions of NOx,VOC, SO2, and CO taken from the underlying gridaccording to national emission inventories. Tra-jectories are calculated every six hours to 740 arrival points covering the whole of Europe, fortime periods of up to six months, using a polar-stereographic grid with a size of 150 ¥ 150km at60°N. This grid size is satisfactory for quantifyingthe transboundary exchange of O3. However, it actually hides large “within-square” variability.This has led to the development of the EMEPEulerian photochemistry model, which has beendesigned as an extension of the EMEP Eulerian aciddeposition model. The grid scale in this model ison a much higher resolution of 50 ¥ 50km. In gen-eral, both modeling approaches are able to repro-duce the broad features of O3 episodes, such as themaximum concentrations and regional distribu-tion. However, comparison between daily maxi-mum observed O3 concentrations for selectedmonitoring stations and the model output from

the Eulerian model generally show a better correla-tion (Simpson et al. 1998). Plate 14.1 (facing p. 180)shows a comparison between the Lagrangian andEulerian models for AOT 40 values generated overthe period April–September 1996. While the spa-tial resolution is clearly different, the models alsodiffer in other respects, such as initial boundaryconditions and the amount of chemistry actuallydetailed.

14.4 DEPOSITION OF NITROGENAND SULFUR ACROSS EUROPE:

ACIDIFICATION ANDEUTROPHICATION

The term “acid rain” gained wide use during the 1970s when the “acidification” of freshwaterlakes in Scandinavia, upland waters in the UK andAdirondack lakes in the northeastern USA werelinked to deposition of acidic species released fromurban/industrial areas. The major component tothis acid precipitation was sulfate (SO4

2-), released

HONO

NO

PAN

Deposition

HO2NO2

RO2NO2

NO2

RO2

O3 HO2

RO2

ROHO

HO NO2

HO2HO HO

RH

NO3

O3

RCO3

N2O5

H2O

HONO2

RONO2

Deposition

Nitrateaerosol

?hv

hv hv

D

D

D

D

Fig. 14.5 A schematic illustratingthe complex chemistry of nitrogenoxides in the atmosphere. (FromSturges 1991, fig. 6.2; with kindpermission from Kluwer AcademicPublishers.)


initially as SO2 from combustion sources usingfuels containing relatively high sulfur content (i.e.coal and oil). The contribution from NOxto overallacidity has in the past been less than that of sulfurfor Europe and North America. However, with the decline of sulfur emissions, NOx has becomethe dominant contributor in relative terms to acidification.

14.4.1 Atmospheric behavior of sulfur and nitrogen

In the atmosphere, SO2, NOx, and NH3 undergo(further) oxidation, ultimately to form sulfuric andnitric acids. SO2 may be oxidized directly by theOH radical to form sulfur trioxide (via an adductthat reacts with O2), which in turn rapidly reactswith water to form sulfuric acid. This is outlined inreactions 14.6–14.8.

(14.6)

(14.7)

(14.8)

Importantly, SO2 is also soluble in water, existingin the aqueous phase as the bisulfite ion (HSO3

-).Aqueous-phase oxidation occurs through severalroutes, including reaction with dissolved O3 or,more importantly, via hydrogen peroxide (H2O2),which is readily soluble in water. Oxidation of SO2in the aqueous phase is an important process with-in cloud droplets and it is estimated that between48 and 84% of all SO2 conversion to H2SO4 inthe troposphere occurs within clouds (Langner &Rodhe 1991). The rate of oxidation of SO2 by thesevarious processes shows a wide variation. For ex-ample, the gas-phase oxidation via OH is on theorder of 10 days, time for significant transportaway from the initial source areas. However, thelifetime of SO2 is actually on the order of a day orso, once other oxidation processes are taken intoaccount, as well as removal via wet and dry deposi-tion. Nevertheless, the formation of oxidationproducts such as sulfate is sufficiently slow toallow for wide atmospheric dispersal, with subse-

SO H O H SO3 2 2 4+ Æ

HSO O SO HO3 2 3 2+ Æ +

SO OH HSO2 3+ Æ

quent deposition of sulfate species occurring over awide regional area.

Nitrogen oxide chemistry is complex and hasbeen highlighted in Section 14.3. Figure 14.5 pres-ents a variety of reactions that NOx may undergo inthe atmosphere. Many of these reactions are re-versible, unlike those of sulfur, with the reactionsdependent on temperature, sunlight, and oxidantconcentrations, changes to which may favor the re-verse of a reaction and formation of the parent NOxagain. Figure 14.5 is best interpreted by focusing on the three rows. The top row compounds are theinorganic nitrates formed from reactions of themiddle row with O3 and the reactive OH and HO2radicals. The bottom row comprises the generic or-ganic nitrates, formed from the reaction of the mid-dle row NOxwith alkyl radicals such as RO and RO2(produced via the reaction of hydrocarbons withHOx). The most common and widely studied com-pound of the organic nitrates is PAN, although theoccurrence and levels of other organic nitrates inboth the polluted and unpolluted atmosphere are ofgrowing interest (Fischer et al. 2000; Kastler et al.2000). The formation of HNO3 can occur directlyvia reaction of NO2 with OH during daylight, orthrough the reaction of NO2with O3at night to formnitrogen trioxide (NO3), which in turn can reactwith NO2 to form nitrogen pentoxide (N2O5). Thisultimately reacts with water to produce HNO3. Theatmospheric lifetime of NOx through these reac-tions and via deposition is approximately 1–2 days,again allowing sufficient time for significant trans-port on a regional basis.

14.4.2 Contemporary emissions of sulfur and nitrogen

Within Europe the emissions of sulfur compoundsand their ensuing deposition is dominated by anthropogenic sources. These emission sourcesare well characterized, with the signatory coun-tries on the various UNECE Protocols producing ayearly emission inventory for the various pollu-tant types (http://projects.dnmi.no/~emep/index.html). Emissions of sulfur (as SO2) across Europe(“European Community”) were estimated to be~12Mt for 1995, compared to a similar amount for


nitrogen (as NO2). Naturally occurring sulfur com-pounds, notably dimethyl sulfide (DMS) emittedfrom oceans, contribute significantly to the globalemission inventory, but apart from coastal areasare not considered to be significant for Europe.Emissions of nitrogen are dominated by both NOx,emitted via traffic and stationary sources, and am-monia (NH3), originating from livestock manureand its handling. Combined emissions of NO2and NH3 greatly outweigh the emissions of SO2. Inaddition, significant amounts of NO may be re-leased from agricultural land, although detailedemissions for the whole of Europe are not yet avail-able, and may be a significant source of NOx inagricultural areas (Skiba et al. 1992). Estimatedemissions of NH3 are shown for the whole of Europe in Plate 14.2 (facing p. 180). These are basedon the EMEP 50 ¥ 50km grid scale and are estim-ated for 1999. Release of reduced nitrogen such asNH3 contributes to eutrophication as well as aciddeposition. Emission estimates are therefore re-quired in order to implement the critical load con-cept (outlined in the next section), based aroundthe chemical buffering capacity of a catchmentsoil.

14.4.3 Acidification and eutrophication

It is now well known that deposition of protons, inassociation with sulfur and nitrogen species, hasdamaged ecosystems across Europe and the north-eastern USA. The first effects of acid precipitationwere observed on aquatic systems, resulting in thedecline of natural fish stocks. The effect of loweredpH in catchment waters is the mobilization ofmetal ions from soil/rock strata, with particularconcern over aluminum, which may be toxic tocertain fish species at lowered pH. In addition, crit-ical nutrients such as phosphate (PO4

2-) (which isfound at trace levels naturally) may be lost as theyprecipitate out with aluminum ions entering alake system. The overall effect of acidification ofnatural waters is a reduction in species richness.Not only are certain fish species vulnerable butalso amphibians and invertebrates. Loss of the latter may even affect bird populations as they

migrate to new areas from which to find an insectfood source.

Damage to vegetation via acidic species is also well established and has been reviewed byWellburn (1994). Direct effects on the foliar partsof vegetation have been observed under both simu-lated and real acid rain conditions, where impact of rainwater with lowered pH results in visibledamage to the leaves of sensitive crop types, suchas radishes, beets, and soya beans. The degree of injury is dependent on many factors, notably theconcentration of acidity in the rain and the dura-tion of contact, but it is also dependent on the temperature, humidity, wind turbulence, and leafmorphology. The effect of acidity can be exacerbat-ed through the evaporation of water from retaineddroplets on the leaf surface, resulting in concen-trating the acidity within the water drop and pro-moting damage to the waxy cuticle of the leaf(Unsworth 1984). On conifer needles cracking ofthe thin waxy plugs covering stomata may occur,allowing ingress of pollutants and pathogens intothe leaf and water loss, and increasing the effects offrost damage.

It is unlikely, however, that these effects alonecan be the cause of the large-scale declines ob-served in forested regions in both Europe (Krause etal. 1986) and North America (Johnson et al. 1986).Acid deposition can also disrupt the soil nutrientstatus due to changes in soil chemistry. Many ofthe essential elements or nutrients, particularlythe cations of potassium (K+), calcium (Ca2+), andmagnesium (Mg2+), may be leached out of the soilvia ion exchange, through increasing inputs of hy-drogen ions (H+). In effect, continual input of aciddeposition reduces the buffering capacity of a soil,although this is highly dependent on the soil typeand the water percolation rate. In a naturally acidicforest soil with low cation content, a large loss of nutrient ions may occur even though the pHchange is minimal. Importantly, as the bufferingcapacity of the soil in the form of available Ca2+

and Mg2+ becomes exhausted, this may result inthe release of Al3+ ions in exchange for H+ ions(Matzner & Prenzel 1992). This mobilization ofAl3+ ions has been suggested as toxic to the uptake


mechanisms of fine root hairs and to the fungi inmycorrhizal association around the roots. Thisprocess of catchment degradation by increasedacidic input through sulfur and nitrogen may con-tribute to forest decline, as well as a decline inother habitat types.

The shift away from sulfur in air pollutionemissions to nitrogen-based pollutants, such asNOx and NH3, has led to increased deposition(both wet and dry) of nitrogen species onto sensi-tive ecosystems. High emissions of NH3 areevident in areas with intensive animal husbandry,such as parts of Belgium and the Netherlands (seePlate 14.2, facing p. 180). Inorganic nitrogen is an essential nutrient for plant growth. However, the increased loading of nitrogen through atmos-pheric deposition can be detrimental to sensitivehabitats. As a consequence, detrimental effectscan occur on the major species comprising thathabitat. A good example of this is presented byWellburn (1994) for a typical forest system and isoutlined here. Initially, increased nitrogen loadingto a forest may result in a growth spurt, usually reflected in an increase in foliar vegetation, butwhich subsequently leads to an imbalance be-tween proteins and carbohydrates. Valuable min-erals such as K+, Mg2+, and PO4

2- may be utilized inorder to address this imbalance, whereby excessnitrogen can be lost as a range of toxic by-products,including amines, amides, and ammonium deriva-tives. These in turn can promote parasitic attackon the tree surfaces. Translocation of carbohy-drates from the roots to other parts of the tree willhave the effect of disrupting the symbiosis thatoften exists between the roots and specific speciesof fungi, resulting in reduced microbial activity inthe soil. The diminishing effects on the roots andenhanced foliar vegetation (leaves) may render thetrees more susceptible to wind-blow damage or excess water loss, effectively increasing the overallstress on the tree.

The inputs of excess nitrogen through deposi-tion associated with regional air pollution may beleading to eutrophication within various ecosys-tems, such as grasslands (Wilson et al. 1995) andcoastal waters (Jaworski et al. 1997), and is a grow-

ing area of concern. A comprehensive review bySmith et al. (1999) presents the impacts of eutro-phication on freshwater, marine, and terrestrialecosystems.

14.4.4 Critical loads and atmospheric deposition

The critical load concept is based on the limits oftolerance that a habitat can have to a certain pollu-tant type —above this limit substantial ecologicalharm could occur. Critical loads are essentially es-timates of the quantities of pollutants that a habi-tat or ecosystem can absorb without ecologicalharm. The concept has been developed with acidi-fication in mind, and limits of tolerance for sulfurand nitrogen acidity have been estimated for mostnatural and seminatural areas within Europe,based on soil quality (acid neutralizing capacity),vegetation type, etc. The 1994 UNECE Protocol onFurther Reduction in Sulfur Emissions (see Table14.1) incorporated the critical load concept as thebasis for setting a new reduction target (Jenkins etal. 1998). In order for critical loads to be used, targetloads need to be set for different areas in order to tryto halt the acidification process. Therefore, targetloads can be either higher or lower than the criticalload values and may be determined by politicalagreement, taking into account socioeconomicfactors as well as scientific findings. A review ofthe calculation procedure for target loads in theNetherlands has been presented by Van der Salmand de Vries (2001). They investigate soil qualitycriteria used to assess critical loads to Dutchforests, and investigate the effect of introducing a second criterion on the existing critical load calculations.

Application of critical loads (for either acidifica-tion or eutrophication) requires accurate assess-ments of pollutant deposition in order to estimatethe exceedance of these loads for a certain area.Models such as the Regional Acidification Infor-mation and Simulation (RAINS) model can predictexceedances and allow the effect of emission re-duction strategies to be assessed for a target area(Amann & Klaassen 1995). However, local varia-


tions in deposition may have a large effect on theaccuracy of such models. This has been investi-gated by examining spatial variability in deposi-tion across an EMEP grid square (150 ¥ 150km),through both statistical evaluation and comparingmodel results with actual deposition measure-ments. From their findings, Hirst et al. (2000) esti-mate the exceedance of acidification critical loadsin Europe to be approximately double those estimated when local variation in deposition is ignored.

Understanding the deposition of sulfur and nitrogen species has therefore been at the forefrontof scientific efforts to investigate the effects of regional air pollution. Importantly, it is necessaryto define the removal rates of pollutants from theatmosphere in order to predict deposition (or more appropriately transboundary exchange) forboth nitrogen and sulfur species. For Europe themodeling of the relationships between emissions(sources) and deposition (receptors) has been car-ried out under the EMEP framework, effectively to assess the impact of policy aimed at curbing pollutant emissions. Both Lagrangian and Eulerianmodels have been developed for the quantitativedescription of the distribution and deposition ofsulfur and nitrogen. A recent example is EURAD(European Acid Deposition Model), an Eulerianmodel intended for the simulation of acidification,nutrification as well as photooxidants over Europe(Borrell et al. 1997). The model encompasses thewhole troposphere and the lower stratosphere andcovers the whole of Europe in a grid of 80 ¥ 80kmor less. The model utilizes a sophisticated meteor-ological model taking meteorological data fromthe European Center for Medium Range WeatherForecasting. In addition, EURAD can provide theboundary conditions on which to run a nestedmodel, the latter operating at much finer spatialresolutions. A nested model is one that runs insideanother, and a good example of this is the EuropeanModeling of Atmospheric Constituents (EUMAC)Zooming model (EZM). This details the mete-orological and photochemical phenomena on anurban scale with grids as small as 5 ¥ 5km and can operate as part of EURAD (Moussiopoulos1994).

This type of approach to modeling, i.e. sett-ing the boundary conditions using a larger-scalemodel such as EURAD and a finer-scale model todepict local-scale processes, has also proved usefulfor the modeling of deposition to sensitive recep-tors (i.e. certain habitats), where assessments ofcritical loads is crucial. One such approach hasbeen utilized in the Netherlands to assess the drydeposition of SO2, using a local-scale depositionmodel incorporated in an EMEP long-range trans-port model (Erisman & Baldocchi 1994). Figure14.6 illustrates this scheme for calculating fine-scale deposition fluxes using a combination ofnested models.

14.5 ARCTIC HAZE

So far this chapter has focused on the industrial-ized regions of Europe and North America. Obvi-ously these regions contain major sources of airpollution and the effects of these pollutants on a regional scale have been observed and reasonablywell characterized. By no means are other parts ofthe globe free from regional air pollution problems:one only has to think of the forest fires of Indonesiain 1997, or the growing number of reports of regional-scale air pollution to come out of the Peo-ple’s Republic of China (e.g. Zhang et al. 1998; Tao& Feng 2000). However, an area that generally re-ceives less attention is the Arctic. This region ex-periences the haze phenomenon; that is, pollutionthat occurs in the lower atmosphere of the Arctic,which builds up during the winter months and isusually observed during March/April after polarsunrise. This has been the subject of scientificstudy over the past 20 years or so and dispels themyth that the Arctic is a pristine area beyond thereach of industrialized pollution. It is now well es-tablished that intrusion of pollutants to this regionoccurs on a seasonal basis, with polluted air beingcarried into the Arctic via long-range atmospherictransport (LRT) from southerly source regions.Aside from the aesthetic impact of Arctic haze, thedeposition of pollutants and their subsequent effects on the fragile polar ecosystem are an area of real concern. This has resulted in international


programs to investigate and understand pollutantimpact on the ecosystem, the foodchain, and ulti-mately the indigenous Arctic peoples (e.g. AMAP1997).

Unlike the Antarctic, the Arctic is entirely surrounded by industrialized regions with highpopulation densities and associated air pollutionproblems. Furthermore, the northern polar regionalso contains its own industrialized areas, notablyin Fenno-Scandinavia as well as Russia, and sup-

ports indigenous peoples who rely on its varioushabitats for their survival and culture. Point pollu-tion events, such as hydrogen bomb tests in theRussian Arctic (1960s), oil spills, and mining activ-ity, are ongoing and fairly frequent events. How-ever, much of the pollution to impact the Arctic is believed to be atmospherically derived throughLRT (AMAP 1997; CACAR 1997). The term “Arc-tic haze” was coined in the 1950s by US pilots, whonoted a brown haze at different layers in the lower

Concentration map

Measurements +

Concentration(evaluation/merge)

Land-useroughness lengthmeteo par: u, Q, rh,Surface wetness

Measurements

Dry depositionmodule

Ra, Rb, RcVd

Subgrid(>5 ¥ 5 km)

50 ¥ 50 km

Time

Time

for evaluation

5 ¥ 5 to 50 ¥ 50 km dry deposition maps

Emission maps<50 ¥ 50 km

Assessment

Daily/seasonal variations Deposition mapsdry, wet,

cloud/fogtotal

European scenarios

Daily/seasonal variations

+

Long-rangetransport model

Fig. 14.6 Flow diagram depictingthe incorporation of a highresolution model within a coarsespatial model, resulting in thecalculation of dry deposition of SO2to simple surfaces. Note that Ra, Rb,Rc relate to “resistances” based onempirical data for depositionmodeling to different surfaces, i.e.vegetation. (From Erisman &Baldocchi 1994; © MunksgaardInternational Publishers Ltd,Copenhagen.)


troposphere during polar sunrise (Mitchell 1956;Raatz 1984). This problem was already known to native people, who called it “poo-jok” or darkhaze. Subsequent scientific investigation revealedthat the haze was derived through transport of airborne pollutants from source regions furthersouth, particularly Eurasia (Rahn & McCaffrey1980). The haze consists of combustion productssuch as carbonaceous particulate matter, heavymetals, and basic air pollutants such as SOx andNOx. Its sources and composition have been thesubject of a review by Barrie (1986).

14.5.1 Pollutant transport into the Arctic and spatial distribution

The prevailing near-surface meteorology duringthe Arctic winter is largely to account for the hazephenomenon. During winter in the northernhemisphere, near surface Arctic meteorology isdominated by four semipermanent atmosphericpressure systems: two centers of low pressure, situated over the Atlantic and Pacific oceans (Icelandic and Aleutian lows), and two high-pres-sure systems, situated over the North Americanand Eurasian landmasses (Raatz 1991). Duringwinter the high-pressure systems develop andbroaden, resulting in intense airflow between themid-latitudes and the Arctic. In the summer thesepressure systems diminish, with a correspondingreduction of airflow into the Arctic. To illustratethis, Fig. 14.7 displays a stylized, typical wintersurface-pressure map for the northern hemi-sphere, highlighting the four dominant pressurecells, along with the accompanying airflow move-ments. Note that the air mass circulation of theAsiatic “high” and the Icelandic “low” comple-ment each other, resulting in enhanced air massmovement into the Arctic and subsequent transferof airborne pollutants from mid-latitudinalsources. Due to these well developed pressure sys-tems, the wintertime circulation in the Arctic andthe mid-latitudes is intense, whereas during thesummer, due to weakening pressure gradients, thenear-surface circulation is considerably slower.The occurrence of pollutants in the Arctic air massis brought about by “episodic” air mass movement

favored by the prevailing meteorology outlinedabove. These air masses carry pollutant burdensthat are representative of the sources over whichthe air parcel passed and result in the rapid and effi-cient transfer of both gaseous and particle pollu-tant loads to the Arctic.

An important feature of the lower Arctic tropo-sphere is the existence of intense temperature inversions. These persist over the whole Arctic region during the winter and over snow- and ice-covered surfaces during the summer. The increasein temperature with height effectively serves as abarrier to vertical turbulent mixing for pollutantsthat enter the Arctic. In winter, surface inversionsform during anticyclonic conditions as a result ofenergy loss from the snow-covered surfaces in theabsence of incoming solar radiation. Typically,surface inversions may extend up to 2000m ac-companied by a temperature increase of ~10°Cfrom the ground to the inversion top. These inver-sions may affect the distribution of pollutants in aparticular air mass. For example, in a well mixedair mass, pollutants in the upper layers may be-come decoupled from the pollutants in the lowerinversion layer, which are then subject to differentremoval rates such as dry deposition. This can re-sult in the occurrence of pollutant “bands” in thelower troposphere. Pollutant bands may also be en-hanced by the occurrence of lifted inversions.These occur when a turbulent mixing layer sepa-rates the inversion base from the ground. The oc-currence of lifted inversions, however, is morefrequent during the summer than the winter,when the Arctic region is impacted by warm moistair.

Probably the most important factor influencingArctic haze is the slow removal rates of pollutantsonce transported into the Arctic. The lack of lightover the winter months, coupled to the lack of precipitation and strong temperature inversions,reduces the dispersal of pollutants, resulting inlong atmospheric lifetimes of air pollutants in thisregion. The spatial distribution of Arctic haze iswidespread across the whole Arctic region, withlevels of SO4

2- (the major aerosol component) gen-erally showing higher levels on the Eurasian side of the Arctic and concentrations declining toward


the high Arctic and northern Canada. A minimumor “saddle” is observed in the region of the Greenland–North American Arctic, which fallsbetween the influence of the Eurasian and easternNorth American sources. Increased precipitationin this area due to the storminess of the easternCanadian and Greenland Arctic also has the effectof decreasing the pollutant loading. Figure 14.8shows the spatial distribution of the arithmeticmean SO4

2- air concentrations (mgm-3) for the period January–April 1980. Superimposed are thethree main sources of air for the North Americanairshed.

14.5.2 Physical and chemical properties ofArctic haze

Arctic haze is actually a mixture of gases andaerosol, although most of the research has focusedon the aerosol fraction due to its role in reducingvisibility. The following information is summa-rized from Sturges (1991) and describes the typical“haze” composition. Aerosol particles are mostnumerous in the diameter range (Dp) of 0.005–0.2mm, with the total number concentration rang-ing from ~10 to 4000cm-3 and a geometric averageof 200–300cm-3. The shape of the number size dis-

IcelandicLow

AsiaticHigh

North Pole

H

H

L

L

Fig. 14.7 Location of the major pressure cells, typical of the lower atmosphere during the Arctic winter.


BC

A

E

M

1.22.1

2.0

2.040°

60°

5

1.1

0.70

0.95

2.0

2.3

2.2

1.8

3.180°

2.04.0

6.0

8.8

8.8

3.2

3.9

6.6

8.0

6.0

4.O

2.0

8.0

60°

0°

60°

180°

120°

3.9

120°

Fig. 14.8 A stylized diagram representing the three major sources of air for the North American airshed (Arctic,Pacific, Caribbean), including the spatial distribution of SO4

2- air concentrations (mgm-3) for the period January–April.(From Sturges 1991, fig. 6.5; with kind permission from Kluwer Academic Publishers.)


tribution is highly variable below 0.1mm, largelydue to the presence (or absence) of nucleationmode particles formed from the condensation ofreactive gases such as SO2. Most of the aerosol oc-curs in the accumulation mode size range (Dp 0.1–1mm) responsible for light scattering and the relat-ed haze. Above 0.2mm the number concentrationdecreases rapidly with increasing diameter, al-though the mass (or volume) of aerosol during thewinter is actually concentrated in the larger parti-cles (Dp 0.1–0.2mm), with a significant contribu-tion from the coarse particle mode (Dp 1–10mm).The number concentration for this size range,however, is approximately four orders of magni-tude less than the accumulation mode. The pres-ence of the giant particle mode (Dp > 10mm) in hazeaerosol has also been reported. Both the coarse andgiant particle modes are believed to originate fromaeolian crustal material such as soil and clay min-erals, as well as sea salt to a lesser extent.

The major component of haze aerosol is SO42-

and to a lesser degree ammonium bisulfate(NH4HSO4). Most of the sulfur (~75%) enters theArctic as SO2, while the remainder (~25%) is al-ready present as the oxidized SO4

2-. This is impor-tant with respect to the nature of the pollution, as oxidation of SO2 to SO4

2- requires oxidants inthe form of OH for gas-phase oxidation or H2O2 orO3 for dissolved aqueous-phase oxidation. Theseoxidants, however, are largely absent during theArctic winter due to the lack of sunlight drivingphotochemical processes, with a result that theratio of SO2/SO4

2- remains high throughout thewinter. Following polar sunrise and the increase inincident sunlight, concentrations of SO2 decreasemore rapidly than those of SO4

-2. Gas-phase oxida-tion of SO2 to SO4

2- during the spring months hasbeen inferred through numerous measurements,where high SO2 levels have been associated withnucleation mode particles, derived through the ox-idation of SO2 to H2SO4. The average compositionof the soluble fraction of aerosol (both fine andcoarse particle sizes) associated with the haze isdisplayed in Fig. 14.9. Non-sea-salt sulfate and itsassociated ammonium ion dominate the mass offine particles, while sea salt (Na+, Cl-) dominates

the coarser-size particles (particularly at Arcticcoastal sites). The sulfates contribute significantlyto the accumulation mode particles that are in-volved in the scattering of sunlight and hence thereduction in visibility resulting in haziness.

Ratios of SO42- with trace metals such as vana-

dium (V) have allowed comparisons to other conti-nental aerosols to distinguish source types andregions. Barrie and Hoff (1984) used the ratios ofSO4

2-/V and SO42-/SO2 in both Arctic and mid-

latitude source regions as an input to a simple

Cl– (18.7%)

Cl– (26%)

Br– (1.5%)

Br– (1%)

Ac– (13.0%)

Ac– (16%)

Fo– (6.2%)

Na+ (9.2%)

Na+ (17%)

K+ (4.2%)

K+ (4%)

NO– (2.2%)3

Fo– (4%)

NO– (3%)3

NH+ (16.5%)4

NH+ (12%)4

SO2–

(28.6%)4

SO2–

(18%)4

Fine

Coarse

Fig. 14.9 Average composition of coarse (2–10mm) andfine (<2mm) aerosols at Barrow, Alaska, in March andApril 1986. (From Sturges 1991, fig. 6.7; with kindpermission from Kluwer Academic Publishers.)


largely through deposition to Arctic surfacesrather than by degradation in the atmosphere, re-sulting ultimately in the exposure of Arctic floraand fauna to these toxicologically relevant contaminants.

14.5.3 Deposition

Deposition of airborne pollutants in the Arctic has been assessed largely through snow pack andice-core sampling. Direct atmospheric sampling of precipitation is difficult, particularly during the winter, due to the low levels of precipitation(especially for the high Arctic) and the problems as-sociated with blowing snow. There have been nu-merous snow pack chemistry surveys conductedin different regions of the Arctic, which have beenreviewed by Barrie (1986). These earlier surveys re-vealed that surface winter snow in many parts ofthe Arctic has a relatively low pH (4.5–5.6), withthe predominant anions being SO4

2- and NO3-.

The snow pack is therefore slightly acidic, withmost of the acidity accounted for by SO4

2- andNO3

-. In an earlier study by Rahn and McCaffery(1979), the concentrations of various trace ele-ments were determined in the winter snowpack of northern Alaska. It was found that the snowpack was enriched in zinc (Zn), cadmium (Cd), andmercury (Hg) relative to the crustal material andthat anthropogenic sources were most likely to account for this.

Ice cores have also demonstrated the impor-tance of the atmospheric deposition of anthro-pogenic pollution. Cores taken on the Agassiz icecap on Ellesmere Island in the Canadian Arcticshow that the conductivity and acidity of the iceare well correlated, and undergo a strong seasonalvariation that mirrors that of Arctic air pollution(e.g. Barrie 1985; Peters et al. 1995). Depositiontrends of PAHs taken from the Greenland ice caphave shown a dramatic increase in concentrationover the past 100 years that correlates well withworld petroleum production. PAH concentrationsrelative to plant fatty acids (C20–C32) demonstratethat contributions of anthropogenic PAHs havesignificantly increased since the 1930s (Kawamuraet al. 1994).

chemical transport model and found that the mean oxidation rate of SO2 between sources andthe Canadian Arctic was 0.04–0.1%h-1 duringthe darkest part of winter (December to February)and increased to 0.2%h-1 in April after polar sunrise.

Reactive nitrate (NO3-) makes a far smaller

contribution to haze aerosol than sulfate, and on a molar basis can be an order of magnitude lowerthan concentrations of sulfur oxides. A seasonaltrend is apparent that is similar to that of sulfate,with high levels during the winter, although a sim-ilar spring maximum does not occur. Importantly,much of the inorganic NOx is actually present asNOy, particularly at higher altitudes, with a largepercentage of oxidized nitrogen occurring as or-ganic forms such as alkyl nitrates, peroxy-nitrateand PAN.

Carbonaceous material is also a major compo-nent of Arctic haze. This has been reported as soot,elemental carbon, or graphitic carbon, althoughthese labels imply some structural form to the car-bon. Therefore, it is more appropriate to term thiscarbonaceous material simply as “black carbon.”Like many of the other haze constituents, levelsare highest during the winter months and closelymatch those of SO4

2-, with a peak in concen-trations occurring between January and March(Hopper et al. 1994). Good correlations have been observed between black carbon and CO2 atseveral locations in the Arctic, indicating the in-fluence of combustion sources on haze pollution.Combustion-derived trace organic compoundshave also been observed, such as the polycyclicaromatic hydrocarbons (PAHs). These semi-volatile persistent organic pollutants have beenobserved in both gaseous and particle phases dur-ing the Arctic winter (Halsall et al. 1997). Exami-nation of the reactive/stable isomers for selectedPAHs (i.e. B[a]P/B[e]P) reveal a ratio that is close tounity throughout much of the winter season, aratio that is not dissimilar to urban observations.This reveals that depletion of the reactive isomer(through either OH attack or some particle sur-face-mediated process) is greatly reduced duringthe dark winter period. Removal from the Arctic atmosphere is therefore likely to occur


14.6 CURRENT TRENDS ANDUNCERTAINTIES IN REGIONAL

AIR POLLUTION

Scientific advances with respect to air pollution,particularly in predictive modeling, have clearlyshaped recent UNECE Protocols as well as na-tional strategies for improving air quality (the twobeing closely linked). Importantly, monitoringnetworks are now well established in both Europeand North America, with 10-year or longer datasets existing for many air pollutants. Indeed, thelist of pollutants has also increased, with VOCsnow included to an extent in the EMEP program, aswell as a selection of persistent organic pollutants.For Europe, long-term monitoring data have beensubject to trend analysis. For O3 there appears to beboth decreasing and increasing trends dependingon location. For example, in the Netherlands adownward trend of 0.4–1.6% per year was evidentfor nine sites between 1981 and 1994, which is in contrast to the other continental EMEP sites(PORG 1997). The monitoring station at Preilla,Lithuania, actually shows a clear upward trend of2.6% per year between 1982 and 1993. This appar-ent difference in trends for different parts of Europemay be down to several factors. Since the 1980scontrols have been implemented to reduce the O3precursors of VOC and NOx (see Table 14.1). Thesecontrols vary from country to country, but forsome countries it is estimated that there have beensignificant reductions in these precursors that willinfluence peak O3 levels (Berge et al. 1994). Thecontrols on anthropogenic releases of VOCs arelikely to have been more effective than those onNOx, affecting the VOC/NOx ratio in certain partsof Europe and hence the rate of O3 formation. Furthermore, from study of trends in O3 and theirprecursors, it is also evident that large-scale pre-vailing meteorology plays a major role in influenc-ing the year-on-year trend. A poor summer typifiedby cloudy conditions and relatively low tempera-tures will result in lower average O3 concentra-tions than a warmer summer (PORG 1997).Generally, however, peak ozone concentrations inEurope are declining, while long-term average con-centrations are increasing.

Importantly, in order to assess the future trendsfor O3, emission inventories for precursors need tobe improved, particularly on the spatial scales onwhich photooxidant models can operate. Speci-ated VOC inventories are clearly lacking over mostof Europe and the quantitative role played by bio-genic VOCs such as isoprene and monoterpenes onO3 yield is an ongoing area of research. Detailedemissions inventories for the other regional pollu-tants outlined in this chapter (such as NH3) are alsorequired to assess ongoing acidification and nutrifi-cation. At the other end of the scale, deposition esti-mates at the fine spatial scale need to be improved,particularly with respect to the critical loads con-cept. An essential prerequisite for this approach isthat the pollutant load (deposition) can be quanti-fied with accuracy and over a sufficient resolutionin both time and space. This is particularly impor-tant with respect to model development for dealingwith complex terrain such as hills and forests,which have a significant influence on the deposi-tion of airborne pollutants.

Finally, it is only controls implemented to abateregional air pollution in temperate latitudes thatwill result in the decline of air pollution in remotelocations such as the Arctic. The focus for this region has been to quantify the contribution of various air pollutants from source areas such asRussia. In a recent study, potential source contri-bution functions (PSCF) have been developed bycombining aerosol data with air parcel back trajec-tories to identify source areas and preferred trans-port pathways. For the winter, high PSCF valuesfor black carbon as well as for the aerosol light scattering coefficient (l = 450nm) were related to industrial sectors in Eurasia (Polissar et al.2001).

REFERENCES

Amann, M. & Klaassen, G. (1995) Cost-effective strate-gies for reducing nitrogen deposition in Europe. Jour-nal of Environmental Management 43, 289–311.

AMAP (1997) Arctic Pollution Issues: A State of the Arctic Environment Report. Arctic Monitoring andAssessment Programme, Oslo.


Barrie, L.A. (1985) Atmospheric particles: their physi-cal/chemical characteristics and deposition processesrelevant to the chemical composition to glaciers. Annals of Glaciology 7, 100–8.

Barrie, L.A. (1986) Arctic air pollution: an overview ofcurrent knowledge. Atmospheric Environment 20,643–63.

Barrie, L.A. & Hoff, R.M. (1984) The oxidation rate and resident time of sulphur dioxide in the arctic atmosphere. Atmospheric Environment 18, 2711–22.

Berge, E., Styve, H. & Simpson, D. (1994) Status of theEmission Data at MSC-W. EMEP MSCW Report 2/95.The Norwegian Meteorological Institute, Oslo.

Borrell, P., Builtjes, P.J.H., Grennfelt, P. & Hov, O. (eds)(1997) Photo-oxidants, Acidification and Tools: PolicyApplications of EUROTRAC Results. The Report ofthe EUROTRAC Application Project. Springer-Verlag,Berlin.

CACAR (1997) Canadian Arctic Contaminants Assess-ment Report. Northern Contaminants Program. De-partment of Indian and Northern Affairs, Ottawa.

Cox, R.A., Eggleton, A.E.J., Derwent, R.G., Lovelock, J.E.& Pack, D.H. (1975) Long-range transport of photo-chemical ozone in north-western Europe. Nature 255,118–21.

Derwent, R.G., Jenkin, M.E. & Saunders, S.M. (1996)Photochemical ozone creation potentials for a largenumber of reactive hydrocarbons under European con-ditions. Atmospheric Environment 30, 189–200.

Dollard, G., Fowler, D., Smith, R.I., Hjellbrekke, A.-G., Uhse, K. & Wallasch, M. (1995) Ozone mea-surements in Europe. Water, Soil and Air Pollution 85,1949–54.

Erisman, J.W. & Baldocchi, M. (1994) Modeling dry depo-sition of SO2. Tellus 46B, 159–71.

Fischer, R.G., Kastler, J. & Ballschmiter, K. (2000) Levelsand patterns of alkyl nitrates and halocarbons in the airover the Atlantic Ocean. Journal of Geophysical Re-search 105, 14,473–94.

Fowler, D., Cape, J.N., Coyle, M. et al. (1999) Modelingphotochemical oxidant formation, transport, deposi-tion and exposure of terrestrial ecosystems. Environ-mental Pollution 100, 43–55.

Fuhrer, J. (1996) The critical levels for effects of ozone oncrops, and the transfer to mapping. In Karenkampi, L.& Skarby, L. (eds.) Critical Levels for Ozone in Europe:Testing and Finalising the Concept. University of Kuopio, Kuopio.

Grunhage, L., Jager, H.J., Haenel, H.D., Lopmeier, F.J. &Hanewald, K. (1999) The European crtical levels for

ozone: improving their usage. Environmental Pollu-tion 105, 163–73.

Halsall, C.J., Barrie, L.A., Fellin, P. et al. (1997) Spatialand temporal variation of polycyclic aromatic hydro-carbons in the Arctic atmosphere. Environmental Sci-ence and Technology 31, 3593–9.

Herman, F., Smidt, S., Huber, S., Englisch, M. &Knoflacher, M. (2001) Evaluation of pollution-relatedstress factors for forest ecosystems in central Europe.Environmental Science and Pollution Research 8,231–42.

Heck, W.W., Taylor, O.C. & Tingey, D.T. (1988) Assess-ment of Crop Loss from Air Pollutants. Elsevier, NewYork.

Hirst, D., Kåresen, K., Høst, G. & Posch, M. (2000) Esti-mating the exceedance of critical loads in Europe byconsidering local variability in deposition. Atmos-pheric Environment 34, 3789–800.

Hopper, J.F., Worthy, D.E.J., Barrie, L.A. & Trivett, N.B.A.(1994) Atmospheric observations of aerosol black car-bon, carbon dioxide and methane in the high Arctic.Atmospheric Environment 28, 3047–54.

Isaksen, I.S.A., Hov, O. & Hesstvedt, E. (1978) Ozone gen-eration over rural areas. Environmental Science andTechnology 12, 1279–84.

Jaworski, N.A., Howarth, R.W. & Hetling, L.J. (1997) Atmospheric deposition of nitrogen oxides onto thelandscape contributes to coastal eutrophication in thenorth-east United States. Environmental Science andTechnology 31, 1995–2004.

Jenkins, A., Helliwell, R.C., Swingewood, P.J., Sefton, C.,Renshaw, M. & Ferrier, R.C. (1998) Will reduced sul-phur emissions under the second sulphur protocol leadto recovery of acid sensitive sites in the UK? Environ-mental Pollution 99, 309–18.

Johnson, A.H., Friedland, A.J. & Dushoff, J.G. (1986) Recent and historic red spruce mortality: evidence ofclimatic influence. Water, Air and Soil Pollution30, 319–30.

Kastler, J., Jarman, W. & Ballschmiter, K. (2000) Multifunctional organic nitrates as constituents in European and US urban photo-smog. FreseniusJournal of Analytical Chemistry 368, 244–9.

Kawamura, K., Suzuki, I., Fujii, Y. & Watanabe, O. (1994)Ice core record of polycyclic aromatic hydrocarbonsover the last 400 years. Naturwissenschaften81, 501–5.

Kley, D., Kleinmann, M., Sanderman, H. & Krupa, S.(1999) Photochemical oxidants: the state of the sci-ence. Environmental Pollution 100, 19–42.

Krause, G.H.M., Arndt, U., Brandt, C.J., Bucher, J., Kenk,G. & Matzner, E. (1986) Forest decline in Europe:


development and possible causes. Water, Air and SoilPollution 31, 647–88.

Krupa, S.V. & Kickert, R.N. (1989) Ambient ozone (O3)and adverse crop response. Environmental Reviews 5,55–77.

Krupa, S.V., Tonneijck, A.E.G. & Manning, W.J. (1989)Ozone. In Flagler, R.B. (ed.), Recognition of Air Pollu-tion Injury to Vegetation: A Pictorial Atlas. Air andWaste Management Association, Pittsburgh.

Langner, J. & Rodhe, H. (1991) A global three-dimensional model of the tropospheric sulfur cycle.Journal of Atmospheric Chemistry 13, 255–63.

Larsen, R.I., McDonnell, W.F., Horstmann, D.H. &Folinsbee, L.J. (1991) An air quality data analysis sys-tem for interrelating effects, standards and neededsource reductions, part II. A log-normal model relatinghuman lung function decrease to O3 exposure. Journalof Air and Waste Management Association 41,455–9.

Lee, E.H., Tingey, D.T. & Hogsett, W.T. (1988) Evaluationof ozone exposure indices in exposure–response mod-elling. Environmental Pollution 54, 43–62.

Liu, S.C., Trainer, M., Fehsenfeld, F.C. et al. (1987) Ozoneproduction in the rural troposphere and the implica-tions for regional and global ozone distribution. Jour-nal of Geophysical Research 92, 4191–207.

Manning, W.J. & Krupa, S.V. (1992) Experimentalmethodology for studying the effects of crops and trees.In Lefohn, A.S. (ed.), Surface Level Ozone Exposuresand Their Effects on Vegetation. Lewis Publishers,Chelsea, MI.

Matzner, E. & Prenzel, J. (1992) Acid deposition in theGerman Solling area: effects on soil solution chemistryand Al mobilization. Water, Air and Soil Pollution 61,221–34.

Mitchell, M. (1956) Visual range in the polar regions withparticular reference to the Alaskan Arctic. Journal of Atmospheric Physics (special supplement), 195–211.

Moussiopoulos, N. (1994) The EUMAC Zooming Model:Model Structure and Applications. EUROTRAC ISS,Garmisch-Partenkirchen.

Oehme, F.W., Coppock, R.W., Mostrom, M.S. & Khan,A.A. (1996) A review of the toxicology of air pollutants:toxicology of chemical mixtures. Veterinary andHuman Toxicology 38, 371–7.

Peters A.J., Gregor, D.J., Teixeira, C.F., Jones, N.P. &Spencer, C. (1995) The recent depositional trend ofpolycyclic aromatic hydrocarbons and elemental car-bon to the Agassiz Ice Cap, Ellesmere Isand, Canada.Science of the Total Environment 160/1, 267–79.

Polissar, A.V., Hopke, P.K. & Harris, J.M. (2001) Sourceregions for atmospheric aerosol measured at Barrow,Alaska. Environmental Science and Technology 35,4214–26.

PORG (1993) Ozone in the United Kingdom. UnitedKingdom Photochemical Oxidants Review Group. Department of the Environment, London.

PORG (1997) Ozone in the United Kingdom. UnitedKingdom Photochemical Oxidants Review Group. De-partment of the Environment Transport and Regions,London.

Raatz, W.E. (1984) Observations of “arctic haze” duringthe “Ptarmigan” weather reconnaissance flights,1948–1961. Tellus 36B, 126–36.

Raatz, W.E. (1991). The climatology and meteorology ofArctic air pollution. In Sturges, W.T. (ed.), Pollution ofthe Arctic Atmosphere. Elsevier Science, Harlow.

Rahn, K.A. & McCaffrey, R.J. (1979) Compositional dif-ferences between arctic aerosol and snow Nature 280,479–80.

Rahn, K.A. & McCaffrey, R.J. (1980) On the origin andtransport of the winter Arctic aerosol. Annals of theNew York Academy of Sciences 338, 486–503.

Schwartz, J. (1989) Lung function and chronic exposureto air pollution: a cross sectional analysis of NHANESII. Environmental Research 50, 309–21.

Skiba, U., Hargreaves, K.J., Smith, K.A. & Fowler, D.(1992) Fluxes of nitric and nitrous oxides from agricul-tural soils in a cool temperate climate. AtmosphericEnvironment 26A, 2477–85.

Simpson, D., Altenstedt, J. & Hjellbrekke, A.G. (1998)The lagrangian oxidant model: status and multi-annual evaluation. In Transboundary Photo-oxidantAir Pollution in Europe: Calculations of TroposphericOzone and Comparison with Observations (EMEPMSC-W Status Report No. 5–29). The Norwegian Meteorological Institute, Oslo.

Smith, V.H., Tilman, G.D. & Nekola, J.C. (1999) Eutroph-ication: impacts of excess nutrient inputs on freshwa-ter, marine and terrestrial ecosystems. EnvironmentalPollution 100, 179–96.

Sturges, W.T. (ed.) (1991) Pollution of the Arctic Atmos-phere. Elsevier Science, London.

Tao, F.I. & Feng, Z.W. (2000) Critical loads of SO2 dry dep-osition and their exceedence in South China. Water,Air and Soil Pollution 124, 499–538.

Unsworth, M.H. (1984) Evaporation from forests in cloudenhances the effects of acid deposition. Nature 312,262–4.

Van der Heyden, D., Skelly, J., Innes, J. et al. (2001) Ozoneexposure thresholds and foliar injury on forest plants


in Switzerland. Environmental Pollution 111, 321–31.

Van der Salm, C. & de Vries, W. (2001) A review of the cal-culation procedure for critical acid loads for terrestrialecosystems. Science of the Total Environment 271,11–25.

Vestreng, V. (2001) Emission data reported to UN-ECE/EMEP: evaluation of the spatial distribution ofemissions. EMEP/MSC-W. Note 1/01. ISSN 0332-9879.

Wellburn, A. (1994) Air Pollution and Climate Change:The Biological Impact, 2nd edn. Longman Scientific &Technical, New York.

White, W.H., Patterson, D.E. & Wilson, W.E. Jr (1983)Urban exports to the nonurban troposphere: Resultsfrom Project MISTT. Journal of Geophysical Research88, 10,745–52.

Wilson, E.J., Wells, T.C.E. & Sparks, T.H. (1995) Are cal-careous grasslands in the UK under threat from nitro-gen deposition. An experimental determination of acritical load? Journal of Ecology 83, 823–32.

Zhang, Y., Shao, K., Tang, X. & Li, J. (1998) The study ofurban photochemical smog pollution in China. ActaScientiarum Naturalium Universitatis Pekinensis 34,392–400.

Documents

Handbook of Atmospheric Science || Regional-Scale Pollution Problems