impact of roadside tree lines on indoor concentrations of tra … · 2014. 4. 24. · impact of...

Impact of Roadside Tree Lines on Indoor Concentrations of Traffic-Derived Particulate MatterBarbara A. Maher,* Imad A. M. Ahmed, Brian Davison, Vassil Karloukovski, and Robert Clarke

Centre for Environmental Magnetism & Palaeomagnetism, Lancaster Environment Centre, Lancaster University, Lancaster LA1 4YQ,United Kingdom

*S Supporting Information

ABSTRACT: Exposure to airborne particulate pollution is associated with premature mortality and a range of inflammatoryillnesses, linked to toxic components within the particulate matter (PM) assemblage. The effectiveness of trees in reducing urbanPM10 concentrations is intensely debated. Modeling studies indicate PM10 reductions from as low as 1% to as high as ∼60%.Empirical data, especially at the local scale, are rare. Here, we use conventional PM10 monitoring along with novel, inexpensivemagnetic measurements of television screen swabs to measure changes in PM10 concentrations inside a row of roadside houses,after temporarily installing a curbside line of young birch trees. Independently, the two approaches identify >50% reductions inmeasured PM levels inside those houses screened by the temporary tree line. Electron microscopy analyses show that leaf-captured PM is concentrated in agglomerations around leaf hairs and within the leaf microtopography. Iron-rich, ultrafine,spherical particles, probably combustion-derived, are abundant, form a particular hazard to health, and likely contribute much ofthe measured magnetic remanences. Leaf magnetic measurements show that PM capture occurs on both the road-proximal and-distal sides of the trees. The efficacy of roadside trees for mitigation of PM health hazard might be seriously underestimated insome current atmospheric models.

■ INTRODUCTION

Human exposure to urban airborne particulate matter (PM) isassociated with a range of damaging health impacts, encompass-ing cardiovascular, respiratory, and neurological damage,1,2 withno known “safe” level.3,4 Depending on its sources, urban PM canconsist of a complex mixture of phases, compositions, and sizes.Particles less than 10 μm in aerodynamic diameter (PM10) canenter the airways and upper lung area, particles of <2.5 μm(PM2.5) can reach the alveolar region, and ultrafine particles(<0.1 μm) enter the blood circulation system.5 Whereas short-term (<1-h) exposure to peak PM10 concentrations has beenassociated particularly with cardiovascular and respiratoryeffects,6,7 longer-term, cumulative exposure to relatively smallincreases (<10 μg m−3) in ambient PM10 and PM2.5

concentrations also impacts adversely on human health.8 Activereduction of ambient, rather than peak, PM10 concentrations hasbeen shown to result in marked improvements in health and life-expectancy outcomes.9 Although the mass of much ambient PMmight consist of compounds of low or negligible toxicity

(whereas metals and particulate-bound polyaromatic hydro-carbons, for example, might specifically contribute to thepathological response), legislative limits for ambient PM10 andPM2.5 concentrations continue to be revised downward.

10 Urbandwellers spend the majority of their time (∼60−90%) indoors,11

where PM can derive from both indoor and outdoor sources.12 Inthe urban environment, traffic can generate large proportions ofindoor PM, as evidenced by the fact that indoor PM peaks havebeen found to closely follow the daily cycles of rush-hour trafficpeaks.13 Whereas PM emissions from industrial sources havedeclined in some areas,14 vehicle-derived PM has increased,because of greater numbers of vehicles and journeys andincreasing proportions of diesel vehicles.15 Primary vehicle-derived PM is emitted directly from vehicle combustion16 and

Received: October 2, 2013Revised: November 8, 2013Accepted: November 11, 2013Published: November 11, 2013

Article

pubs.acs.org/est

© 2013 American Chemical Society 13737 dx.doi.org/10.1021/es404363m | Environ. Sci. Technol. 2013, 47, 13737−13744

pubs.acs.org/est

abrasion processes, especially brake wear.17 Vehicle movementscan also cause PM to be resuspended from the road surface.Secondary PM forms in the atmosphere through chemicalreactions and condensation.Although it is well-known that urban trees can reduce ambient

outdoor PM concentrations by capturing particles on theiraerodynamically rough leaf surfaces,18−20 the efficacy of urbantrees for PM reduction is intensely debated. City- and street-scalemodeling studies suggest reductions in PM10 of as little as <1%and as much as ∼60%, respectively.21,22 Modeling also suggeststhat dense tree crowns can worsen in-canyon PM concentrations,by reducing dispersion and atmospheric mixing.23−26 Measure-ments of PM10 reduction by trees remain sparse,18 especially inurban areas and at the local scale. Critical in terms of PMreduction are the velocities of particle deposition onto treeleaves. Reported values vary by 3 orders of magnitude, from∼0.01 to 10 cm s−1,27 depending not only on particlecharacteristics and meteorology but on tree species, leafarchitecture, and surface properties.28,29 Mitchell et al.,28 forexample, reported (magnetic) PM10 deposition velocities of ∼5cm s−1 for birch, a species that also displays low emissionpotentials for volatile organic compounds (VOCs).23 Carefulselection, planting, and maintenance of tree species with suchhigh deposition velocities and low VOC emissions mightpotentially offer an efficient and cost-effective means of removingPM from the atmosphere and improving human health outcomesworldwide. Improvement of air quality is particularly importantin street canyons, abundant within city environments, wherelowered wind speeds, in-canyon air recirculation, and high trafficemissions can result in frequent exceedances of PM10 stand-ards.12

Here, we examine the impact of trees on indoor air quality atthe local scale, by temporarily installing a line of young trees(silver birch) outside a row of terraced houses adjacent to aheavily trafficked urban road, in Lancaster, northwest England.Indoor PM concentrations were measured over a ∼2-weekperiod, using a combination of novel magnetic measurements ofTV screen wipes in all of the houses, and PMmonitoring inside a“control” house and a second house, inside the artificial tree line.Leaf PM capture sites and PM size, shape, and composition wereexamined by scanning electron microscopy (SEM). Our resultsdemonstrate the effectiveness of carefully selected tree lines insubstantially reducing indoor PM10 concentrations, throughparticle capture and agglomeration on tree leaves, and the valueof fast, inexpensive magnetic measurements in quantifying spatialdifferences in PM, particularly iron-rich, ultrafine, combustion-derived PM.

■ EXPERIMENTAL METHODSIndoor PM concentrations were measured for a row of eightterraced houses, adjacent to a heavily trafficked (∼12000 vehiclesper day, ∼80% cars), uphill section of road, in the city ofLancaster, northwest England (Figure 1). All of the houses are ofsimilar age, have double-glazed windows and central heating withexternal flues, and are occupied by nonsmokers.Prior to the experiment, indoor PM concentrations were

monitored in the ground-floor, road-adjacent rooms of twohouses (house 25, May 9−11, 2013, and house 31, Mar 27−Apr18 and May 9−11, 2013) using two intercalibrated Grimm dustmonitors (model 1.106, Labortechnik Ltd., Ainring, Germany).These monitors are designed for real-time counting,30 as time-weighted averages every 10 min, of volumetric particleconcentrations, with particle size discrimination into 15 size

ranges, from >0.3 to 20 μm. House 25 is occupied by a family offour; the front room is a formal dining room. In house 31, thefront room is a study bedroom of a multiple-occupancy studenthouse. No cooking, smoking, fires, or any other major indoor PMsources existed in either of the monitored rooms, nor were thereany changes in the normal use or activity levels within the twomonitored, closed rooms before and after tree installation.Thirty young birch trees were then installed at the curbside

outside four of the houses (Figure 1, houses 31, 33, 35, and 37)for a period of 13 days (May 11−23, 2013). Indoor PMconcentrations were monitored, post-tree installation, in house25 (non-tree-lined, NTL) and house 31 (tree-lined, TL).Meteorological data (temperature, wind speed, direction, rain,

and humidity) were obtained for the duration of the experimentfrom the Hazelrigg Weather Station, Lancaster University [seeFigure S1, Supporting Information (SI)]. Traffic flow data werekindly provided by Lancaster City Council, for the period Apr23−29, 2013 (Figure S2, SI).A novel, fast, and inexpensive method was used to measure the

integrated record of indoor PM deposition within each of theeight houses, by measuring the saturation isothermal remanentmagnetization, in an applied dc field of 0.3 T (SIRM0.3T) of amoist paper swab taken from the screen of the household’stelevision or an installed computer monitor (see SI forinstrumentation details). All screens [plasma or light-emittingdiode (LED)] displayed insignificant levels of surface charge,whether switched on or off. Both one week before and again atthe start of the 13-day experimental period, each available TV ormonitor screen was cleaned, and at the end of the monitoringinterval, a constant screen area was swabbed (using a woodensampling frame) with a readily available “wet wipe”. Themagnetic remanence of the screen swabs was corrected bysubtraction of the wet wipe “blank”. Similar magnetic remanencemeasurements were made of samples of the birch leaves (fiveleaves/sample) from the road-proximal and -distal sides of thetree line. The leaf SIRMs (in units of A m2) were normalized forthe adaxial (upper) surface area (m2) of the young leaves, whichranged from ∼7 to 16 cm2.

Figure 1. (a) Terraced row of houses, numbered (house 37 is only partlyvisible at the right-hand side), and (b) temporary installation of roadsidebirch trees.

Environmental Science & Technology Article

dx.doi.org/10.1021/es404363m | Environ. Sci. Technol. 2013, 47, 13737−1374413738

SEM analyses were performed of clean (pre-experiment) birchleaves and of eight birch leaves at the end of their roadsideexposure, using a Leo Gemini 1530 field-emission-gun- (FEG-)SEM instrument (Cambridge, U.K.) equipped with an OxfordInstruments energy-dispersive X-ray (EDX) detector, with high-resolution capability down to ∼200 nm and reproducibleaccuracy to approximately ±3%. Leaf disks were cut on a cleansurface and mounted on a carbon double-sided tape fixed on asupport stub; degassed (for 1 h at 0.7 bar); coated with a 10-nmPt/Pd film; and imaged at 5, 10, or 15 kV. SEM-energy-dispersiveX-ray analysis (EDXA) was used to identify the elementalcompositions of a range of particles of different sizes andmorphologies. To reduce detection levels below the typical limit(∼1000 ppm by weight), we allowed for a long counting time(∼2 min). To avoid Al/Cu signal overlap from the support stuband sample, a clean glass disk was fixed between the carbon tapeand the stub.

■ RESULTS AND DISCUSSION

Differences in indoor PM10 concentrations in the two monitoredhouses can, of course, reflect not only the impact of the tree linebut also any differences in indoor or outdoor PM sources, indooractivity levels, air exchange rates, meteorological conditions,and/or reduced data scatter as the total data count increased.However, both before and during the experiment, measured

peaks in indoor PM concentrations in houses 25 and 31 occurredrather synchronously and at regular intervals during eachweekday, namely, at ∼7:30−9:30 a.m., 12:00−2:00 p.m., 3:00−4:00 p.m., and 4:30−7:30 p.m. (Figure 2), that is, associated withpeak traffic flows (Figure S2, SI). Exceptions to this pattern wereevident on Friday, Mar 29, 2013 (a public holiday in the U.K.),when these peaks were absent (Figure S3, SI), and duringweekends, when the first PM10 peak occurred later, from ∼9:00or 9:30 a.m., and peak times varied throughout the weekend(Figure 2). The synchroneity of indoor PM peaks in the two

Figure 2. PM10 concentrations inside TL house 31 during the periods (a) Apr 11−16, 2013 (pre-tree-line installation), and (b)May 16−21, 2013 (post-tree-line installation). The weekend PM10 peaks (light shading) are lower and start later in the morning than the weekday patterns. Note the reductionsin PM10 after installation of the tree line.

Figure 3. PM10 concentrations inside NTL house 25, pre- and post-tree installation (on May 11, 2013). Weekend PM10 data are shaded.



houses and the absence/reduction of peaks during the publicholiday indicate that ambient, traffic-derived PM is the majorcontributor to the indoor PM concentrations in the monitoredrooms.In both monitored houses, indoor PM10 values varied from <1

to >50 μg m−3 but concentrated within the 1−10 μg m−3 range(Figures 2 and 3). Construction work at NTL house 27,involving replacement of mortar, might have been responsible forepisodic “spikes” in indoor PM concentrations (>50 μg m−3) inNTL house 25 (Figure 3).TL house 31 displays substantial reductions in indoor PM10

values in the post-tree installation interval (Figure 2). In contrast,for NTL house 25, little change in PM10 values is evident fromthe pre- to post-tree installation period (Figure 3). Noting thateach data point represents a measurement of indoor PM at 10-

min intervals and that we seek to quantify the changes in ambientPM during the experiment, we use density distribution plots tomap the PM1, PM2.5, and PM10 data sets (Figure 4). It is clearthat, post-tree installation, each PM size fraction declines inconcentration inside TL house 31, both in absolute terms and inrelation toNTL house 25. For example, pre-installation,∼70% ofthe TL house 31 data comprise PM10 concentrations rangingfrom 0.5 to 5.8 μg m−3, compared with a markedly reduced rangeof 0.5−2 μg m−3 after tree installation. This PM10 reduction inTL house 31 has no comparable counterpart in NTL house 25;the house 25/house 31 PM ratio rises, post-tree installation, foreach PM size fraction (Figure 4a−f).Although PM monitoring was possible for only two of the

houses, fast, inexpensive, and sensitive magnetic analysis of theTV screen wipes provides quantified information on the

Figure 4. Two-dimensional contour plots of indoor PM concentrations in TL house 31 and PM ratio for NTL house 25/TL house 31: (a,c,e) pre- and(b,d,f) post-installation of tree line. The color coding represents the likelihood of a PM value falling within a particular concentration (with a probabilityscale ranging from 0 to 1).



integrated PM deposition for all of the houses. Prior to theexperiment, no systematic difference was evident in the TV wipeSIRM0.3T values available from house 23 to house 39 (Figure 5a).At the end of the experiment (Figure 5b), for each of the TL

houses, the SIRM0.3T values for the front-room TV screen swabsare lower than those from the NTL houses by between 52% and64% (average 60%). These results are remarkably consistent withthe Grimm PM data, which independently identify >50%

Figure 5. SIRM0.3T values for TV/monitor screen wipes, showing the concentrations of deposited magnetic particles, for (a) the pre-experimentalperiod, with no trees installed, and (b) the post-experimental period, for non-tree-lined houses and tree-lined houses (marked with *). 25M identifies theSIRM0.3T value for the occupants’ own TV, located in a middle room; 25F is for a computer screen placed in the front room.

Figure 6. (a−d) Scanning electron micrographs of the adaxial surface of the exposed birch leaves contaminated with PM. (e,f) EDXA spectra ofdeposited particles shown in micrographs c and d.



reductions in indoor PM10, PM2.5, and PM1 concentrations,following installation of the tree line (Figure 4). Although theabsolute SIRM0.3T values differ substantially between the NTLand TL houses, the ranges in the sets of measured values aresimilar (∼1.5 × 10−8 and ∼1.25 × 10−8 A, respectively).Two screen swabs were obtained from NTL house 25; the

SIRM0.3T value from the household’s own television (located in amiddle ground-floor room) was 50% lower than that from acomputer monitor placed in the front (road-proximal) room(Figure 5b).The magnetic particle loadings on the birch tree leaves

increased substantially after their exposure at the roadside. TheSIRM0.3T values of the road-proximal leaves fell in the range of(5−14)× 10−6 A (average 7.6× 10−6 A, 83% higher than the pre-exposed leaf SIRM0.3T value of 4.1× 10−6 A), and for thepavement-side leaves, SIRM0.3T = (4.4 to 12) × 10−6 A (average7.0 × 10−6 A, 71% higher).Scanning electron micrographs (Figure 6a,b) show the typical

rough, hairy morphology of the adaxial leaf surfaces of theselected silver birch species (Betula utilis Doorenbos), which arehypostomatic (i.e., stomata are present only on the underside ofthe leaves). After their roadside exposure, the leaves displayeddeposited PM, occurring as distinct clusters and agglomerations(∼103 particles, Figure 6c), particularly in proximity to leaf hairs(Figure 6b) and trapped inside the microindentations (height ×width ≈ 4 × 3 μm) of the rough leaf surface (Figure 6d).The leaf-deposited PM comprised particles with a wide range

of diameters up to∼30 μm, but based on SEM image analysis, themajority (>60%) were submicrometer in diameter (Figure 6c,d).Individual submicrometer particles (<100 nm) often occurred asaggregates forming larger agglomerates of particles (Figure 6e,f).From EDXA, major PM elemental contributions included (>1 wt%) C > O > Fe > Ti > Si > Ca > Al > Na > K > Cl >Mg (Figure 6e,f; also see Figure S4, SI). Many of the coarser particles (∼1−10μm) were irregular in shape and size (Figure 6d), theircomposition dominated by the major elements (Ca, Na, Al, Si,Cl), suggesting some possible contribution from natural sourcessuch as soil dust. In contrast, many of the finer particles (<200nm) exhibited spherical or semispherical morphologies, withsmooth surfaces (Figure 6d), and were particularly rich in Fe(10−55 wt %). Such Fe-rich particles are abundant and typical ofcondensation droplets released from high-temperature combus-tion processes.31 They are likely to contribute much of themeasured magnetic remanence of the TV swabs and the treeleaves. Higher concentrations of magnetic particles have beenreported for PM collected from gasoline engine exhausts thanfrom diesels.32 Nanoscale magnetite particles (<30 nm) appearstrongly associated with exhaust emissions, whereas largermagnetites (≳1 μm) are associated with abrasion of metallicvehicle components, including particles from disk brakes.33,34

The high Fe content (up to ∼55 wt %) of many particles,especially the combustion-derived nanoparticles, might repre-sent a particular potential hazard to health. Such particles arelikely to be biologically active, given their large surface area;enrichment in transition metals; and reported association withparticle-bound organic species, including known carcinogenssuch as benzo[a]pyrene.35 A strong correlation has beenreported for urban dusts between sample magnetic contentand mutagenicity.36 Iron-rich, magnetically detectable particlesmight cause oxidative stress by direct generation of reactiveoxygen species (ROS) not only in lung cells37 but also in thebrain.38 Oxidative brain damage is a characteristic of most types

of neurodegenerative disease, including Alzheimer’s andParkinson’s disease.38,39

Whereas particle deposition rates depend on a range of factors(e.g., wind speed, surface orientation, particle diameter), theinfluence of leaf surface properties on particle deposition,capture, and agglomeration have so far been underparametrizedin atmospheric modeling of urban PM. The efficacy of PMcapture by a single line of young trees was demonstrated here by,first, two independent approaches, conventional PM monitoringand magnetic measurement of screen swabs, both indicating>50% reduction in measured indoor PM concentrations; second,SEM analysis indicating the effective agglomeration of PM(especially submicrometer PM) around leaf hairs and withinregions of negative surface relief; and, third, magnetic remanencemeasurements of the tree leaves identifying similar levels ofmagnetic PM capture by the road-proximal and -distal leaves.All of these data are consistent with the efficient interception

and capture of vehicle-derived PM by the tree line, rather than(and/or as well as) any airflow perturbation and physicalscreening effects and “fumigation” of the roadway. A value of 0.64cm s−1 is frequently used for the deposition velocity of PM10 onleaf surfaces in atmospheric modeling studies,27,40 even in that ofPugh et al.,22 where 60% PM removal was reported (for a greenwall, in a street canyon with low wind speed and long particleresidence time). Because 0.64 cm s−1 appears to be an order ofmagnitude too low, given published data28 and the reduction inindoor PM10 concentrations observed here, Pugh et al.’s modelcould be run with less extreme vegetation, wind, and/orresidence time constraints and still predict ∼60% PM removal.Interception and capture of PM by our selected species, silverbirch, appears to be enhanced by both interfingering of leaf hairsinto the airflow and agglomeration of PM into aggregatescomposed of large numbers (>1000) of particles.In terms of urban greening, the data indicate the potential

effectiveness of the edge effect for PM reduction. Rather thanincreasing total urban tree cover per se, single roadside tree linesof a selected, high-deposition-velocity, PM-tolerant speciesappear to be optimal, preferably managed so as to preventdevelopment of a dense canopy crown and facilitate atmosphericdispersion of PM. Whether PM deposition on leaves continuescumulatively through the in-leaf season41 or equilibrates withambient atmospheric PM concentrations28 appears to depend ontree species, leaf changes through the growing season, and rainfallevents. Rainfall can wash deposited PM off leaves and intodrainage courses and/or the soil.18,29,42,43 Even after leaf fall, thebark and branches of roadside trees intercept PM.44,45 Manage-ment and disposal of PM-contaminated leaf fall should beconsidered within any urban greening strategy.

■ ASSOCIATED CONTENT

*S Supporting InformationTime-series meteorological data (wind speed, temperature, andrain) from Hazelrigg Weather Station, Lancaster, U.K., togetherwith indoor PM10 data; PM10 concentrations vs time during aU.K. national holiday (Good Friday); time-series of traffic counts(as hourly number of vehicles) vs indoor PM10; and acompilation of SEM-EDXA data collected from SEM-analyzedbirch leaves. This material is available free of charge via theInternet at http://pubs.acs.org.



http://pubs.acs.org

■ AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected]; phone: +44-0-1524 510268;fax: +44-0-1524 593985.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

The tree line installation was organised and funded by the BBC.We thank the University of Leeds EPSRC Nanoscience andNanotechnology Facility (LENNF) for access to the SEM/EDAX and Mr Stuart Micklethwaite for his technical help. Wealso thankDr Stuart Lawson (Physics Department) for assistancewith SEM, and the reviewers for their thoughtful comments.

■ REFERENCES(1) Pope, C. A.; Dockery, D. W. Health effects of fine particulate airpollution: Lines that connect. J. Air Waste Manage. Assoc. 2006, 56, 709−742.(2) Nel, A.; Xia, T.; Madler, L.; Li, N. Toxic potential of materials at thenanolevel. Science 2006, 311, 622−627.(3) Brunekreef, B.; Holgate, S. T. Air pollution and health. Lancet 2002,360, 1233−1242.(4) Dominici, F.; Daniels, M.; Zeger, S. L.; Samet, J. M. Air pollutionand mortality: Estimating regional and national dose−responserelationships. J. Am. Stat. Assoc. 2002, 97, 100−111.(5) Pickrell, J. A.; Erickson, L. E.; Klabunde, K. J. Toxicity of inhalednanomaterials. In Nanoscale Materials in Chemistry, 2nd ed.; Klabunde,K. J., Richards, R. M., Eds.; John Wiley & Sons, Inc., 2009; Chapter 22,pp 729−769.(6) Curtis, L.; Rea, W.; Smith-Willis, P.; Fenyves, E.; Pan, Y. Adversehealth effects of outdoor air pollutants. Environ. Int. 2006, 32, 815−830.(7) Schwarze, P. E.; Øvrevik, J.; Lag, M.; Refsnes, M.; Nafstad, P.;Hetland, R. B.; Dybing, E. Particulate matter properties and healtheffects: Consistency of epidemiological and toxicological studies. Hum.Exp. Toxicol. 2006, 25, 559−579.(8) Quantification of the Effects of Air Pollution on Health in the UnitedKingdom; Committee on the Medical Effects of Air Pollutants(COMEAP), Department of Health: London, 1998.(9) Pope, C. A.; Ezzati, M.; Dockery, D. W. Fine-particulate airpollution and life expectancy in the United States. N. Engl. J. Med. 2009,360, 376−386.(10) Integrated Science Assessment for Particulate Matter; Report EPA/600/R-08/139F; U.S. Environmental Protection Agency: Washington,DC, 2009.(11) Adgate, J. .; Ramachandran, G.; Pratt, G. .; Waller, L. .; Sexton, K.Spatial and temporal variability in outdoor, indoor, and personal PM2.5exposure. Atmos. Environ. 2002, 36, 3255−3265.(12) Harrison, R. M.; Thornton, C. A.; Lawrence, R. G.; Mark, D.;Kinnersley, R. P.; Ayres, J. G. Personal exposure monitoring ofparticulate matter, nitrogen dioxide, and carbon monoxide, includingsusceptible groups. Occup. Environ. Med. 2002, 59, 671−679.(13) Diapouli, E.; Chaloulakou, A.; Mihalopoulos, N.; Spyrellis, N.Indoor and outdoor PM mass and number concentrations at schools inthe Athens area. Environ. Monit. Assess. 2008, 136, 13−20.(14) Emissions of Primary Particles and Secondary Particulate MatterPrecursors (CSI 003); European Environment Agency (EEA):Copenhagen, Denmark, 2008.(15) Transport Trends 2007 Edition; HMSO: London, 2007.(16) Harrison, R. M.; Deacon, A. R.; Jones, M. R.; Appleby, R. S.Sources and processes affecting concentrations of PM10 and PM2.5particulate matter in Birmingham (U.K.). Atmos. Environ. 1997, 31,4103−4117.(17) Pant, P.; Harrison, R. M. Estimation of the contribution of roadtraffic emissions to particulate matter concentrations from fieldmeasurements: A review. Atmos. Environ. 2013, 77, 78−97.

(18) Fowler, D.; Skiba, U.; Nemitz, E.; Choubedar, F.; Branford, D.;Donovan, R.; Rowland, P. Measuring aerosol and heavy metaldeposition on urban woodland and grass using inventories of 210Pband metal concentrations in soil. Water Air Soil Pollut. Focus 2004, 4,483−499.(19) Bradshaw, A. D.; Hunt, B.; Walmsley, T. Trees in the UrbanLandscape: Principles and Practice; Taylor & Francis: London, 1995.(20) Beckett, K. P.; Freer-Smith, P. H.; Taylor, G. Urban woodlands:Their role in reducing the effects of particulate pollution. Environ. Pollut.1998, 99, 347−360.(21) Nowak, D. J. Pollution removal by Chicago’s urban forest. InChicago’s Urban Forest Ecosystem: Results of the Chicago Urban ForestClimate Project; McPherson, E. G., Nowak, D. J., Rowntree, R. A., Eds.;General Technical Report NE-186; Northeastern Forest ExperimentStation, Forest Service, USDA: Radnor, PA, 1994; pp 63−81.(22) Pugh, T. A. M.; MacKenzie, A. R.; Whyatt, J. D.; Hewitt, C. N.Effectiveness of green infrastructure for improvement of air quality inurban street canyons. Environ. Sci. Technol. 2012, 46, 7692−7699.(23) Gromke, C.; Ruck, B. On the impact of trees on dispersionprocesses of traffic emissions in street canyons. Boundary-Layer Meteorol.2009, 131, 19−34.(24) Hofman, J.; Stokkaer, I.; Snauwaert, L.; Samson, R. Spatialdistribution assessment of particulate matter in an urban street canyonusing biomagnetic leaf monitoring of tree crown deposited particles.Environ. Pollut. 2013, 183, 123−132.(25) Vos, P. E. J.; Nikolova, I.; Janssen, S. A high-order model foraccurately simulating the size distribution of ultrafine particles in a traffictunnel. Atmos. Environ. 2012, 59, 415−425.(26) Nowak, D. J.; Civerolo, K. L.; Trivikrama Rao, S.; Sistla, G.; Luley,C. J.; Crane, D. E. A modeling study of the impact of urban trees onozone. Atmos. Environ. 2000, 34, 1601−1613.(27) Litschke, T.; Kuttler, W. On the reduction of urban particleconcentration by vegetationA review.Meteorol. Z. 2008, 17, 229−240.(28) Mitchell, R.; Maher, B. A. Evaluation and application ofbiomagnetic monitoring of traffic-derived particulate pollution. Atmos.Environ. 2009, 43, 2095−2103.(29) Beckett, K. P.; Freer-Smith, P. H.; Taylor, G. Particulate pollutioncapture by urban trees: Effect of species and windspeed. Global ChangeBiol. 2000, 6, 995−1003.(30)Manual for the Dust Monitor, 3rd ed.; Grimm Labortechnik, Ltd.:Ainring, Germany, 1996; Sections 1.104−1.106.(31) Petrovsky, E.; Ellwood, B. B.Magnetic monitoring of air, land, andwater pollution. In Quaternary Climates, Environments and Magnetism;Maher, B. A., Thompson, R., Eds.; Cambridge University Press:Cambridge, U.K., 1999; Chapter 8, pp 279−322.(32) Chaparro, M. A. E.; Marie, D. C.; Gogorza, C. S. G.; Navas, A.;Sinito, A. M. Magnetic studies and scanning electron microscopyX-ray energy dispersive spectroscopy analyses of road sediments, soils andvehicle-derived emissions. Stud. Geophys. Geod. 2010, 54, 633−650.(33) Sagnotti, L.; Macrì, P.; Egli, R.; Mondino, M. Magnetic propertiesof atmospheric particulate matter from automatic air sampler stations inLatium (Italy): Toward a definition of magnetic fingerprints for naturaland anthropogenic PM10 sources. J. Geophys. Res. Solid Earth 2006, 111,B12S22.(34) Sagnotti, L.; Taddeucci, J.; Winkler, A.; Cavallo, A. Composi-tional, morphological, and hysteresis characterization of magneticairborne particulate matter in Rome, Italy. Geochem. Geophys. Geosyst.2009, 10, QO8Z06.(35) Halsall, C. J.; Maher, B. A.; Karloukovski, V. V.; Shah, P.; Watkins,S. J. A novel approach to investigating indoor/outdoor pollution links:Combined magnetic and PAHmeasurements. Atmos. Environ. 2008, 42,8902−8909.(36) Morris, W. A.; Versteeg, J. K.; Bryant, D. W.; Legzdins, A. E.;McCarry, B. E.; Marvin, C. H. Preliminary comparisons betweenmutagenicity and magnetic susceptibility of respirable airborneparticulate. Atmos. Environ. 1995, 29, 3441−3450.(37) Borm, P. J. A.; Kelly, F.; Kunzli, N.; Schins, R. P. F.; Donaldson, K.Oxidant generation by particulate matter: From biologically effectivedose to a promising, novel metric.Occup. Environ. Med. 2007, 64, 73−74.



mailto:[email protected]

(38) Allsop, D.; Mayes, J.; Moore, S.; Masad, A.; Tabner, B. J. Metal-dependent generation of reactive oxygen species from amyloid proteinsimplicated in neurodegenerative disease. Biochem. Soc. Trans. 2008, 36,1293−1298.(39) Block, M. L.; Zecca, L.; Hong, J.-S. Microglia-mediatedneurotoxicity: Uncovering the molecular mechanisms. Nat. Rev.Neurosci. 2007, 8, 57−69.(40) Escobedo, F. J.; Nowak, D. J. Spatial heterogeneity and airpollution removal by an urban forest. Landscape Urban Plann. 2009, 90,102−110.(41) Kardel, F.; Wuyts, K.; Maher, B. A.; Samson, R. Intra-urban spatialvariation of magnetic particles: Monitoring via leaf saturation isothermalremanent magnetisation (SIRM). Atmos. Environ. 2012, 55, 111−120.(42) Beckett, K. P.; Freer-Smith, P. H.; Taylor, G. The capture ofparticulate pollution by trees at five contrasting urban sites. Arboric. J.2000, 24, 209−230.(43) Hansard, R.; Maher, B. A.; Kinnersley, R. P. Rapid magneticbiomonitoring and differentiation of atmospheric particulate pollutantsat the roadside and around twomajor industrial sites in the U.K. Environ.Sci. Technol. 2012, 46, 4403−4410.(44) Smith, W. H.; Staskawicz, B. J. Removal of atmospheric particlesby leaves and twigs of urban trees: Some preliminary observations andassessment of research needs. Environ. Manage. 1977, 1, 317−330.(45) Zhang, C.; Huang, B.; Piper, J. D. A.; Luo, R. Biomonitoring ofatmospheric particulate matter using magnetic properties of Salixmatsudana tree ring cores. Sci. Total Environ. 2008, 393, 177−190.



impact of roadside tree lines on indoor concentrations of tra … · 2014. 4. 24. · impact of...

Documents