UBC Faculty Research and Publications

Air quality in postunification Erfurt, East Germany : associating changes in pollutant concentrations… Ebelt, Stefanie; Brauer, Michael; Cyrys, Josef; Tuch, Thomas; Kreyling, Wolfgang G.; Wichmann, H.-Erich; Heinrich, Joachim Apr 30, 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Ebelt_S_et_al_Air_quality_postunification_Erfurt.pdf [ 125.82kB ]
JSON: 52383-1.0074684.json
JSON-LD: 52383-1.0074684-ld.json
RDF/XML (Pretty): 52383-1.0074684-rdf.xml
RDF/JSON: 52383-1.0074684-rdf.json
Turtle: 52383-1.0074684-turtle.txt
N-Triples: 52383-1.0074684-rdf-ntriples.txt
Original Record: 52383-1.0074684-source.json
Full Text

Full Text

The unification of East and West Germanyin 1990 brought major social and politicalchanges, particularly affecting East Germany(1). This restructuring resulted in significantchanges in emissions of air pollutants.During the 1980s, the former GermanDemocratic Republic was a major industrialpower with a high level of manufacturingoutput. This high output was fed by hugeinputs of natural resources, which in turn ledto extreme emissions of pollutants (2). In1990, the European Union (3) concluded,the environment in East Germany is in a cata-strophic state. Water and air pollution is so badthat it is no longer simply a matter of cleaning upthe environment but one of restoring the mostbasic living conditions. Clean up in East Germany initially occurredas a by-product of unification; low demandfor products and energy led to a collapse ofalmost the entire industrial and agriculturalstructure in East Germany, sharply decreas-ing emissions between 1989 and 1991. Thiswas followed by air pollution controlsimplemented between 1992 and 1996 (4,5).A decade has passed since the Germanreunification, and it is timely to examine theresults of these structural changes anddecreased emissions on the air quality. The sharp change in emissions in thepast 10 years has been an opportunity for anatural experiment to evaluate the healthimpacts of air pollution. For example,decreases in total suspended particles (TSP)and sulfur dioxide between 1991 and 1995were associated with decreased prevalence ofinfectious airway diseases in East Germanchildren (6,7). Additionally, Erfurt, a city informer East Germany, was the site of severallarge-scale epidemiologic studies during the1980s (8) and early 1990s (9) concerning thehealth effects of air pollution. Since this time,the composition of Erfurt’s ambient air haschanged, and assessments of the effects ofthese changes on the health status of the for-mer East German population are currently inprogress (10). Current particulate air pollution healtheffects research is focused on evaluating bio-logic mechanisms that may explain the epi-demiologic associations (11). Exposureassessment and analysis of the compositionof ambient particles and the impact of spe-cific sources is of special interest. Most epi-demiologic studies have assessed exposure ofthe study population in terms of particulatemass concentrations (i.e., in micrograms percubic meter), which are measured for spe-cific particle cut sizes. For example, the massconcentrations of particles with aerodynamicdiameters (d a) ≤ 10 µm or 2.5 µm providePM10 or PM2.5 concentrations, respectively.Seaton et al. (12) suggested that ultrafineparticles (particles with diameters < 0.1 µm)could be a component of the ambient parti-cle mixture that is responsible for theobserved health effects and suggested amechanistic hypothesis. Because ultrafineparticles do not contribute significantly tothe total mass of particles, measurementsbased only on mass concentrations do notaccurately represent the concentrations ofultrafine particles; thus ultrafine particleshave been quantified in terms of particlenumber concentrations. Several epidemio-logic studies have evaluated the impacts ofultrafine particle levels. In Erfurt, an increaseEnvironmental Health Perspectives • VOLUME 109 | NUMBER 4 | April 2001 325Air Quality in Postunification Erfurt, East Germany: Associating Changes inPollutant Concentrations with Changes in EmissionsStefanie Ebelt,1 Michael Brauer,1 Josef Cyrys,2 Thomas Tuch,2 Wolfgang G. Kreyling,3 H.-Erich Wichmann,2 and Joachim Heinrich21The University of British Columbia, School of Occupational and Environmental Hygiene, Vancouver, British Columbia, Canada;2GSF–National Research Center for Environment and Health, Institute of Epidemiology, Neuherberg, Germany; 3GSF–National ResearchCenter for Environment and Health, Institute of Inhalation Biology, Neuherberg, GermanyAddress correspondence to M. Brauer, TheUniversity of British Columbia, School ofOccupational and Environmental Hygiene, 2206East Mall, Vancouver, BC V6T 1Z3 Canada.Telephone: (604) 822-9585. Fax: (604) 822-9588.E-mail: brauer@interchange.ubc.caResearch described in this article was supported inpart by the Health Effects Institute (research agree-ment 95-10), an organization jointly funded by theU.S. Environmental Protection Agency and auto-motive manufacturers. Funding was also providedby a Career Investigator Award from the AmericanLung Association and a Scientist Award from theMedical Research Council of Canada and theBritish Columbia Lung Association to M. Brauer.Received 29 June 2000; accepted 17 October2000.ArticlesThe unification of East and West Germany in 1990 resulted in sharp decreases in emissions ofmajor air pollutants. This change in air quality has provided an opportunity for a natural experi-ment to evaluate the health impacts of air pollution. We evaluated airborne particle size distributionand gaseous co-pollutant data collected in Erfurt, Germany, throughout the 1990s and assessed theextent to which the observed changes are associated with changes in the two major emission sources:coal burning for power production and residential heating, and motor vehicles. Continuous data forsulfur dioxide, total suspended particulates (TSP), nitric oxide, carbon monoxide, and meteorologicparameters were available for 1990–1999, and size-selective particle number and mass concentrationmeasurements were made during winters of 1991 and 1998. We used hourly profiles of pollutantsand linear regression analyses, stratified by year, weekday/weekend, and hour, using NO and SO2 asmarkers of traffic- and heating-related combustion sources, respectively, to study the patterns of var-ious particle size fractions. Supplementary data on traffic and heating-related sources were gatheredto support hypotheses linking these sources with observed changes in ambient air pollution levels.Substantially decreased (19–91%) concentrations were observed for all pollutants, with the excep-tion of particles in the 0.01–0.03 µm size range (representing the smallest ultrafine particles thatwere measured). The number concentration for these particles increased by 115% between 1991and 1998. The ratio of these ultrafine particles to TSP also increased by more than 500%, indicat-ing a dramatic change in the size distribution of airborne particles. Analysis of hourly concentrationpatterns indicated that in 1991, concentrations of SO2 and larger particle sizes were related to resi-dential heating with coal. These peaks were no longer evident in 1998 due to decreases in coal con-sumption and consequent decreased emissions of SO2 and larger particles. These decreases in coalcombustion and the decreased concentrations of SO2 and particles of larger size classes may have ledto decreased particle scavenging and may be partially responsible for the observed increases in ultra-fine particles. Traffic-related changes, such as increased numbers of trucks and increased use ofdiesel vehicles in Erfurt, were also associated with increased number concentrations of ultrafine par-ticles. Morning particle peaks of all sizes were associated with NO and CO (markers for traffic) inboth the 1991 and 1998 periods. There were significant differences in the ultrafine particle levels formorning hours between 1991 and 1998, suggesting that traffic was the cause of this increase. Keywords: air pollution, coal combustion, environmental exposure, motor vehicles, particles, sulfurdioxide, ultrafine particles. Environ Health Perspect 109:325–333 (2001). [Online 7 March 2001]http://ehpnet1.niehs.nih.gov/docs/2001/109p325-333ebelt/abstract.htmlin the number of ultrafine particles was asso-ciated with a decrease in peak expiratoryflows (PEF) of asthmatic adults, which wasstronger than the effects of PM10 (13,14). InErfurt, associations have also been foundbetween ultrafine particles and medicationuse by asthmatics and symptom reporting bycoronary heart disease patients (15). A studyin Finland on the PEF of asthmatic childrendid not indicate a stronger association withultrafine particles, but consistent and signifi-cant results were obtained with PM10 andblack smoke exposure metrics (16). From extensive monitoring conducted inErfurt throughout the 1990s, includinggravimetric methods and aerosol size spec-trometry, the concentrations of particles bothin terms of their mass and number have beenfollowed over time. Detailed assessments ofthe size distribution of particles have alsobeen undertaken. For example, during thewinter of 1991–1992, the mean numberconcentration was largely composed of ultra-fine particles (72%), whereas the bulk of theparticle mass (83%) was due to accumulationmode particles, which are in the size range of0.1–0.5 µm (17). Since this time, the massconcentrations of fine particles (PM2.5) havedecreased by 72%; however, the overall num-ber concentrations have remained constant,with particles < 0.03 µm increasing by 82%(18,19). Due to the potential health risksassociated with ultrafine particles, the sourceof the observed increases in recent years is ofimmediate concern. Additionally, identifica-tion of important sources contributing toambient concentrations of ultrafine particlesmay provide useful information to supportthe development of reduction policies in thisand in other locations. Two source categories that are large con-tributors to air pollution and that haveundergone significant change in EastGermany since the unification include energyproduction (power plants and residentialheating) and vehicles (5). As was commonthroughout East Germany before the unifica-tion, lignite (brown coal) was the major fuel,meeting approximately 70% of the region’senergy requirements (4). Lignite has high sul-fur content, and its combustion results inhigh emissions of SO2 and particulate matter(PM) when burned (20). At the time of uni-fication in 1990, lignite-fired power plants inthe former East Germany operated with min-imum emission controls. Since this timethere have been large reductions in emissionsof particles and SO2 from these sources dueto the shutdown of old plants, transitionfrom the use of coal to liquid and gaseousfuels, reduction in the sulfur content of coal,and retrofits of lignite-fired power plantswith flue gas desulfurization systems (4,5,21).West German plants had already achievedemission reduction in the 1980s; thus theunification of East and West Germanybrought together two power-generation sys-tems. Thuringia, the state in which Erfurt islocated, is an industrial region. It still hashigher atmospheric SO2 concentrations rela-tive to other German states (with valuessometimes > 50 µg/m3), but much lowerthan in 1990 or earlier, when annual SO2average concentrations were > 150 µg/m3 (5). Concurrent with the changes in energyproduction, mobile sources have undergonetransitions. Due to the absence of other avail-able vehicles, the car fleet in East Germanywas largely composed of “Trabants” prior tounification. These were small cars with 26horsepower, two-stroke engines and visiblyhigh exhaust emissions (1). After unificationthe fleet was replaced with vehicles producedlargely by Western countries (1), which hadcomparatively modern engine technology,including three-way catalysts. Additionally,during this same period, diesel-powered vehi-cles have become more common throughoutWestern Europe (22,23). Further, mostmotorcycles during the preunification periodwere equipped with two-stroke engines.Significant changes in fuels and combus-tion processes associated with energy produc-tion and motor vehicles in East Germanyhave occurred, and these have likely impactedthe ambient pollutant concentrations overtime. In this study we used a unique databaseof particle size measurements and continuousgaseous air pollutant data from Erfurt to ini-tially describe their temporal patterns duringthe 1990s. Furthermore, we assessed theextent to which the observed changes in parti-cle size distribution are associated withchanges in specific emission sources—namely,energy- and traffic-related combustionsources.MethodsData collection. Erfurt (population approxi-mately 201,100) is a city in the state ofThuringia, Germany, approximately 200 mabove sea level and about 100 km east of theformer east–west border. It lies on a flatplain, surrounded on all sides by a 100-mhigh ridge, except toward the north. Thisgeography favors wintertime inversions,which result in elevated levels of ambient airpollution during this season. Local air pollu-tion sources include motor vehicles, residen-tial heating, small-scale industry, and districtheating plants. Suburban areas contain largeapartment complexes that are heated withsteam supplied by a large coal-burningpower plant located several kilometers north-east of the city center. The inner city has alarge historical center where buildings wereheated by individual coal furnaces until thetime of reunification.Ambient particulate, gaseous, and meteo-rologic data were collected in Erfurt through-out the 1990s and have been used previouslyin numerous epidemiologic and exposureassessment studies (13,17,24–26). Between 1October 1991 and 31 March 1992, andbetween 1 October 1995 and 31 March1999, measurements were made at theInstitute of Hygiene site located approxi-mately 1 km south of the Erfurt city centerand approximately 40 m from the nearestmajor road. The spatial representativeness ofthis site has been analyzed in detail and is gen-erally representative of the air quality withinErfurt (26). In previous work, PM10 and sul-fate measurements from this site were foundto be significantly correlated (Spearman rankcorrelations: 0.69–0.85) with those made atother locations within Erfurt, although somespatial variability was evident during periodsof low wind speed.The number concentrations (NCs) ofambient fine particles were determined witha mobile aerosol spectrometer (MAS) as pre-viously described by Brand et al. (27). TheMAS is a combination of two instrumentsthat measure different size ranges. Particleswith diameters between 0.01 and 0.5 µm arequantified using a differential electricalmobility sizer (DMPS). The DMPS consistsof a differential electrical mobility analyzer(TSI Model 3071; TSI, St. Paul, MN, USA)used to classify the particles by their electri-cal mobility, which are then counted by acondensation particle counter (TSI models3760, 3010). Particles with diametersbetween 0.1 and 2.5 µm are classified withan optical laser aerosol spectrometer (LAS-X;Model LAS-X; PMS Inc., Boulder, CO,USA). The instruments used at the Erfurtsite are described in more detail elsewhere(17,28). Overall, the particle number distri-bution of the ambient aerosol (using the 13different size channels from the DMPS and15 channels for each of 4 size ranges fromthe LAS-X) was measured every 6 min. Themass distribution of particles was determinedfrom the number distribution basing calcula-tions on a mean particle density of 1.5 × 103kg/m3, and an assumption of spherical parti-cles (17). Sulfate concentrations were deter-mined from 24-hr PM2.5 Harvard Impactor(with the addition of a citric acid-coatedhoneycomb denuder for acidity sampling)measurements taken on a daily basis duringwinter 1991 and every second day after1995. Filters were analyzed for SO42- by ionchromatography. We measured nitric oxideseparately from nitrogen dioxide with a two-channel chemiluminescence monitor (mod-els NH3OM, AC31M; Environment S.A.,Poissy Cedex, France). Sulfur dioxide wasmeasured via UV absorption (ModelAF21M; Environment S.A.). Both monitorsArticles • Ebelt et al.326 VOLUME 109 | NUMBER 4 | April 2001 • Environmental Health Perspectivesoperated on a continuous basis, producing3-m averaged measurements. Temperature,relative humidity (Model FR3205-M; RCI,Planegg, Germany) and windspeed sensorswere mounted on the sampling shelter,thereby providing continuous meteorologicdata. Additionally, hourly measurements forSO2, TSP, NO, and CO from as early as1990 was available from governmental mon-itoring stations (Thueringer Landesanstaltfuer Umwelt). Data analysis. The continuous numberconcentration data obtained from theDMPS and LAS-X channels were onlystored for certain size classes, from which wechose three ranges for analysis. The smallestfraction included particles with diametersbetween 0.01 and 0.03 µm (NC0.01–0.03),thus capturing the lowest measurable por-tion of ultrafine particles (ultrafine particlesare usually defined as particles with sizes upto 0.1 µm). The second size class includedparticles with diameters between 0.03 and0.5 µm (NC0.03–0.5), covering the upper por-tion of the ultrafine mode and the lowerportion of accumulation mode particles(which are defined as particles between 0.1and 1.0 µm). The largest size class containedall particles in the size range between 0.5 and2.5 µm diameter (NC0.5–2.5), thus includingthe upper portion of the accumulation modeas well as fine-mode coarse particles. Thesesize class divisions were slightly differentfrom previous reports using these measure-ments, since the continuous data needed forhourly assessments was stored for limitedsize classes only (13,14,17,28). From the calculated mass distribution,PM2.5 concentrations were determined byadding over all particle size ranges between0.01 and 2.5 µm. For all NC, PM2.5, gaseous(NO, NO2, CO, SO2), and meteorologicvariables (temperature, relative humidity,wind speed), the measurements from eachhour were averaged. The resulting hourlymeasurements were then used to obtain 24-hr(daily) averages. For SO42- only daily averageswere available. We used daily averages toassess the distributions of individual pollu-tants as well as their temporal patterns. Formost variables, measurements during the1991 monitoring period were only takenfrom 1 October 1991 to 31 March 1992; nosummer data were collected. In addition,gaseous measurements for January to March1992 were not available. Due to these gaps inthe data, two distinct time periods were cho-sen for further comparative analyses: 3-monthtime windows from October to Decemberwere compared for the years 1991 and 1998.Analyses included t-tests for differences inmean concentrations between these years. Wecreated hourly profiles by plotting the medianvalues for each hour of the day for each timeperiod (for NC, NO, CO, and SO2). Thesedata were stratified by weekday and weekendperiods. To visually compare hourly timeseries for the different pollutants, the hourlymedian values were standardized by calculat-ing their percentage of the daily average con-centration. Hourly data were subsequentlyassessed by linear regression analyses using theunivariate general linear model procedure inSPSS version 9.0 (SPSS, Chicago, IL, USA).We predicted NC fractions using year andArticles • Changes in air quality and emissions in GermanyEnvironmental Health Perspectives • VOLUME 109 | NUMBER 4 | April 2001 3278007006005004003002001000SO2 (µg/m3 )Jan 90May 90Sep 90Jan 91May 91Sep 91Jan 92May 92Sep 92Jan 93May 93Sep 93Jan 94May 94Sep 94Jan 95May 95Sep 95Jan 96May 96Sep 96Jan 97May 97Sep 97Jan 98May 98Sep 98Jan 99May 99Sep 99Figure 1. Daily SO2 time series. Data from the Thueringer Landesanstalt fuerUmwelt monitoring network. 6005004003002001000TSP  (µg/m3 )Jan 90May 90Sep 90Jan 91May 91Sep 91Jan 92May 92Sep 92Jan 93May 93Sep 93Jan 94May 94Sep 94Jan 95May 95Sep 95Jan 96May 96Sep 96Jan 97May 97Sep 97Jan 98May 98Sep 98Jan 99May 99Sep 99Figure 2. Daily TSP time series. Data from the Thueringer Landesanstalt fuerUmwelt monitoring network. 350300250200150100500NO (µg/m3 )May 91Sep 91Jan 92May 92Sep 92Jan 93May 93Sep 93Jan 94May 94Sep 94Jan 95May 95Sep 95Jan 96May 96Sep 96Jan 97May 97Sep 97Jan 98May 98Sep 98Jan 99May 99Sep 99Figure 3. Daily NO time series. Data from the Thueringer Landesanstalt fuerUmwelt monitoring network. 76543210CO (mg/m3 )Sep 91Jan 92May 92Sep 92Jan 93May 93Sep 93Jan 94May 94Sep 94Jan 95May 95Sep 95Jan 96May 96Sep 96Jan 97May 97Sep 97Jan 98May 98Sep 98Jan 99May 99Sep 99Figure 4. Daily CO time series; data from the Thueringer Landesanstalt fuerUmwelt monitoring network. weekend/weekday as fixed factors and NOand SO2 concentrations as covariates.Autocorrelation of measurements over timewas addressed by using the hour of the day asan additional fixed factor. We also investigatedeffects of two-way interactions between theindependent variables. Variables were nottransformed for these analyses. For ease ofcomparison, effects of NO and SO2 wereassessed in the same models, although therewas a relatively high correlation between them(Spearman rank correlation, 0.67); their effectsas covariates in separate models were similar. ResultsTrends over time. Due to incomplete hourlydata for several of the pollutants at varioustimes within the period 1990–1999, wefocused analyses on two shorter periods forwhich complete data were available:October–December of 1991 and 1998.These periods were also selected because theyrepresent the extremes of long-term trends inconcentrations for most pollutants (Figures1–4). SO2 and TSP (Figures 1 and 2) exhib-ited steady declines in concentrationbetween 1991 and 1998, largely due to theminimization of winter extremes. NO andCO (Figures 3 and 4) concentrations alsodecreased, but to a lesser extent. Analyseswere restricted to winter periods because thetypical seasonal pattern in Erfurt exhibitshigh concentrations in winter months andlower concentrations in summer months.Table 1 presents summary statistics foreach variable for 1991 and 1998. The vari-ables were not normally distributed, with theexception of temperature and relativehumidity. NC0.03–0.5, NO, and NO2 wereapproximated by lognormal distributions,whereas all other metrics were not better rep-resented when log transformed. Both arith-metic and geometric descriptive statistics arereported for consistency. Based on compari-son of arithmetic means, concentrationsdecreased significantly for all pollutants (t-test, p ≤ 0.001). For example, we observedreductions of 58% for TSP, 74% for PM2.5,50% for SO42-, and 91% for SO2. In con-trast, NC0.01–0.03 increased by 115%. TheNC0.01–0.03:TSP ratio changed from 78 to394, which is a 5-fold increase from 1991 to1998. The NC0.01–0.03:PM2.5 ratio changedby a factor of 8, from 83 to 686. Theseratios emphasize the drastic shift of theaerosol size distribution toward the smaller,ultrafine particles from 1991 to 1998.Relative humidity was slightly higher in1998 relative to 1991 (p < 0.001), whileambient temperatures were similar duringthe two periods. The mean windspeed waslower in 1998 than in 1991.From these initial observations we devel-oped several hypotheses to evaluate the impactof various sources that have changed overtime. First, the main fuels used for energy pro-duction have changed from brown coal to nat-ural gas. Lowering the PM emissions fromhigh-sulfur coal combustion could lead to thedecreases observed for NC0.03–0.5 andNC0.5–2.5. Such decreases could in turn leadto decreases in particle scavenging, thusallowing the NC0.01–0.03 fraction to increase.Particle scavenging may occur when largenumbers of fine and coarse particles providesurface area onto which ultrafine particles dif-fuse and readily aggregate. Reductions in theemissions of these larger particles couldreduce the available surface area and thereforeincrease the atmospheric residence time ofultrafine particles. Second, there has been aninflux of Western-style vehicles employingmodern engine technology in easternGermany. It has been hypothesized thatimproved efficiencies associated with moderncombustion engines result in more completecombustion and larger proportions of parti-cles < 0.03 µm. Increases in the number ofdiesel vehicles in Erfurt could also contributeto increased ultrafine particle counts.As there was no direct measure forattributing changes in these sources tochanges in ambient particle concentrations,we evaluated the hour-to-hour changes inthe concentrations of gaseous pollutantmarkers in relation to each particle size frac-tion. Gaseous pollutants and particle sizefractions showing similar temporal structureswere assumed to have arisen from the samesources. For our purposes, hourly SO2 datawere used as an indicator for coal-firedpower generation and residential heating,Articles • Ebelt et al.328 VOLUME 109 | NUMBER 4 | April 2001 • Environmental Health PerspectivesTable 1. Descriptive statistics for each pollutant by year. 1998/1991 RatioMeasurement Year No.a AM SD Range GM GSD AM GMNC0.01–0.03 1991 70 7,896 7,252 369–25,657 4,159 4 2.15 3.48(n/cm3) 1998 75 16,982 9,672 4,045–51,620 14,491 2NC0.03–0.5 1991 70 10,593 8,624 1,121–39,073 7,324 3 0.63 0.74(n/cm3) 1998 75 6,669 4,896 1,436–25,177 5,427 2NC0.5–2.5 1991 70 122 128 8–625 78 3 0.20 0.18(n/cm3) 1998 75 25 35 2–186 14 3TSPb 1991 70 101.34 77.76 19.85–421.83 79.68 2.05 0.42 0.47(µg/m3) 1998 75 43.05 28.56 12.00–157.00 37.55 1.73PM2.5 1991 70 95.01 85.57 10.51–402.10 67.85 2.36 0.26 0.29(µg/m3) 1998 75 24.75 21.28 3.74–102.98 19.87 2.02SO42- 1991 87 8.24 6.84 1.01–32.59 7.16 2.06 0.50 0.51(µg/m3) 1998 38 4.11 5.25 0.39–22.29 3.68 2.09NO 1991 69 47.43 47.06 6.38–231.82 32.73 2.45 0.59 0.46(µg/m3) 1998 90 22.75 33.90 0.30–189.09 11.35 3.36NO2 1991 69 39.43 18.34 11.95–106.57 36.40 1.61 0.81 0.81(µg/m3) 1998 90 28.80 16.48 6.35–81.70 25.76 1.74COb 1991 92 1.45 1.09 0.21–6.18 2.27 1.43 0.33 0.63(mg/m3) 1998 92 0.48 0.42 0.10–2.15 1.43 1.27SO2 1991 69 91.23 84.04 6.35–365.15 62.52 2.53 0.09 0.11(µg/m3) 1998 90 6.58 7.32 0.33–50.87 5.63 2.09Temperature 1991 92 4.22 5.04 –7.20–14.30 — — 0.99 —(°C) 1998 90 4.18 5.19 –6.86–15.27 — —RH 1991 92 75 9 55–94 — — 1.17 —(%) 1998 90 88 6 67–100 — —Windspeed 1991 92 2.19 1.23 0.00–5.70 2.96 1.49 0.68 0.79(m/sec) 1998 90 1.49 0.89 0.38–4.83 2.35 1.39Abbreviations: AM, arithmetic mean; GM, geometric mean; GSD, geometric standard deviation; RH, relative humidity. Different numbers of samples were collected for the various particlemeasurements, but for comparison purposes, summaries of NC, TSP, and PM2.5 were based only on days for which each of these measurements were available. aNumber of 24-hr averages out of a possible 92 days for each variable. bData from Thueringer Landesanstalt fuer Umwelt–Kraempferstrasse site; all other measurements were fromInstitute of Hygiene/GSF site. and NO and CO data were used as indica-tors for traffic-related emissions. Temporal patterns in energy production.Ambient sulfur oxide gases are formed largelywhen fuels containing sulfur, such as coaland oil, are burned (20). In East Germanybefore the unification, sulfur-rich coal wastypically used as a source of energy for indus-try and heating purposes (4). Homes wereeither heated by individual coal-burningovens or through long-distance steampipelines fed by coal-fired power plants. SO2was used as a marker of such heating-relatedcombustion.In 1991, morning SO2 peaks were moreprominent during the week compared to theweekend (Figure 5). This pattern is consis-tent with morning peaks being due to energyuse by industry during the workday. EveningSO2 peaks occurred to the same extent onweekends as during the week, and these werethought to be related to home heating.These curves are consistent with typicalhome heating patterns during this period.The 1998 profiles were much weaker, withan altered hourly pattern. No longer weremorning and evening peaks observed; rather,only one prominent and prolonged mid-dayincrease was present. Supplemental data on energy productionwere gathered to help explain the observedchanges in SO2 concentrations. Over the past9 years, the percentages of homes heatedlocally (60%) compared to those heated bylong-distance steam pipelines (40%) did notchange, but the sources of energy used byeither route did (Table 2). In the Erfurt areathere were four power plants in operationbefore unification, three of which were subse-quently shut down between 1990 and 1993;the fourth plant was retrofitted. Between1991 and 1998, the use of coal by powerplants decreased by 84% and was replaced bya 3-fold increase in natural gas consumption.Also, between 1992 and 1995, the use of coalfor residential heating decreased by 45%.Additionally, the sulfur content of coal wasreduced from 2% in 1991 to 1.7% in 1993,after which the percentage remained con-stant. The decreased use of coal and increaseduse of gas can explain the large difference inSO2 concentration between 1991 and 1998in Erfurt. The small SO2 elevations duringthe day in 1998 could be due to the remain-ing coal that was still used at this time or dueto regional transport.Temporal patterns in traffic. NO andCO are important air pollutants associatedwith motor vehicle emissions (20). On-roadmotor vehicles contribute up to 35% of thetotal nitrogen oxides (NOx) emissions in theUnited States, of which more than 95% ofNOx from light-duty, gasoline vehicles isreleased as NO (29). Hourly NO and COprofile plots (Figures 6 and 7) support the useof these pollutants as traffic indicators inErfurt. During the week, morning NO andCO peaks were observed that corresponded tothe morning rush hour. In 1991, these morn-ing NO and CO peaks occurred slightly ear-lier than morning SO2 peaks, supporting thedifference in sources of these pollutants: NOand CO arising from travel to work and SO2being emitted from factories in operation dur-ing the workday. Unfortunately, no trafficcount data were available with which todirectly verify the observations.Peaks in NO and CO concentrations didnot occur to the same extent on weekends,likely due to less traffic on the road. Overtime, the weekday morning traffic peakshifted: in 1998, the peak occurred an hourlater and was longer. This could be due tochanges in work patterns that delay andlengthen the rush hour. Before unification,morning shifts in large factories usuallystarted at 0700 hr. After the shutdown of thelarge plants and the implementation of flexi-ble work hours, traffic was likely shiftedtoward start of regular business hoursaround 0900 hr.We collected supplementary data onvehicles in Erfurt to help explain the reduc-tions in NO and CO concentrations anddifferences in the timing of concentrationpeaks between 1991 and 1998. In Erfurt,the total number of vehicles (cars, motorcy-cles, trucks) over the last 7 years did notchange (Table 3). However, a large increase(268%) in the percentage of trucks and alarge decrease in the number of vehiclesequipped with two-stroke engines, includinga drastic decrease in the number of motorcy-cles (83%), was observed. The fraction oftrucks with diesel motors has been steadilyincreasing in Germany in the past 20 years,from 58% in 1980 to 88% in 1999 (30).Since 1990, the number of diesel trucks inGemany has doubled (30). In terms of cars, the total number hasdoubled in East Germany since 1991 (5). InThuringia, the number also increased by afactor of 1.6 (30), although in Erfurt, thenumber of private cars increased only slightly(15%) between 1990 and 1998 (Table 3).However, the composition of the private carfleet has likely changed dramatically in east-ern Germany due to modern technology anddiesel motors. For example, even though theproportion of diesel cars in Thuringia(7–8%) has been lower than in the country(13%) during the 1990s, the total number ofArticles • Changes in air quality and emissions in GermanyEnvironmental Health Perspectives • VOLUME 109 | NUMBER 4 | April 2001 329Figure 5. SO2 hourly profiles for 1991 and 1998, showing weekends (WE) and weekdays (WD) separately. 100.0090.0080.0070.0060.0050.0040.0030.0020.0010.000.00SO2 (µg/m3 )Hour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241991 WD1998 WD1991 WE1998 WETable 2. Temporal changes in energy sources.Energy source/type Measure 1991 1998 1998/1991 RatioHome heating Percent of homes 40 40 1Steam pipe (long distance)aResidential (total)a Percent of homes 60 60 1Residential (coal)b Percent of homes 60 33 0.55Power plantCoala Tons 565,172 92,243 0.16Oilc Tons 12,620 13,716 1.09Gas m3 × 103 26,492 102,439 3.87Residential refers to heating source present in the residence itself. aData from Schmidt (42). bData from the microcensus in 1992 and 1995 (43). cData from Strom und Fernwaerme GmbH (44).diesel vehicles has more than doubled since1992, increasing from 11% to 16% of thetotal on-road vehicle population (30).Relating emissions to particle levels. Theemissions from heating and traffic (asassessed by SO2 and NO/CO) were relatedto particulate measurements to determinethe effect of these sources on the ambientparticulate mix. Standardized weekdayhourly profile plots (which combined thenumber concentrations of the three sizeclasses of particles and the gases) were cre-ated (Figures 8–11) and regression modelsusing these variables (except CO) were con-structed (Table 4). We conducted visual andstatistical analyses to assess the patterns overtime (between 1991 and 1998), betweenweekdays and weekends and between thehours of a day.From the regression models, theNC0.01–0.03 fraction was significantly lowerin 1991 compared to 1998, as also shown bythe ratio comparisons (Table 1). As indi-cated by the weekend/weekday factor (Table4), all particle fractions were significantlylower on weekends compared to weekdays.This supports traffic and energy production(when combining industry and residentialsectors) as contributors to all size fractions,as they are both expected to be higher duringthe week. Both visual assessments of thehourly profiles and regression models indi-cated significant positive associationsbetween all particle size fractions and both ofthe gaseous pollutants, NO and SO2.Considering interactions between winter andthe gases (winter × NO, winter × SO2), theeffects of both NO and SO2 on NC0.01–0.03were significantly different in 1991 com-pared to 1998. This suggests that the influ-ences of both traffic and energy productionon NC0.01–0.03 have changed over time.Reviewing these results and the literature,there are several reasons to suggest that trafficis the major factor associated with theobserved increase in NC0.01–0.03. First, due totheir tendency to aggregate, ultrafine particlenumber concentrations are more likely to beelevated near sources (31). Considering thelocation of the monitoring stations in Erfurt,the influence of power plants on ultrafineparticles in this study is likely small com-pared to the influence of vehicles. Second,Figures 8 and 10 show that morning NCpeaks of all particle sizes in both years occurearlier than the morning SO2 peak. Thus,particles at this time of day were not likelyassociated with power plant emissions.Instead, the temporal patterns of the curvescoincided well with the curves for NO andCO, suggesting that traffic was associatedwith the morning NC peaks of all size frac-tions. Indeed, from the winter × hour inter-actions, NC0.01–0.03 in 1991 was significantlylower in the morning (approximately0700–1000 hr) compared to the same hoursin 1998. This supports the hypothesis of achange in source for NC0.01–0.03, especiallyduring the morning hours. This is consistentwith motor vehicles as a major source becausemotor vehicles have undergone a changedhourly pattern over time.The hourly profile plots also suggest thathome heating was associated with theevening trends of larger particles. TheNC0.03–0.5 and NC0.5–2.5 fractions followedthe SO2 level with respect to timing in theafternoon. In 1998, there were no promi-nent afternoon peaks for these larger frac-tions or for SO2, which could be expected ifhome heating with coal were much reducedin this year. Therefore, SO2 in 1991 wasfound to be associated with large particle lev-els in the evening, which we believe to bedue to residential coal burning. For both1991 and 1998, all morning particle peakswere associated with NO and CO. Becausewinter × hour interactions indicated thehours of 0700–1000 hr as being significantlydifferent for NC0.01–0.03 between the twoyears, we suggest that changes in motor vehi-cle emissions between 1991 and 1998 (com-position of vehicle fleet, driving patterns,emission controls) have contributed to theincrease in NC0.01–0.03.Articles • Ebelt et al.330 VOLUME 109 | NUMBER 4 | April 2001 • Environmental Health PerspectivesFigure 7. CO hourly profiles for 1991 and 1998, showing weekends (WE) andweekdays (WD) separately. (mg/m3 )Hour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241991 WD1998 WD1991 WE1998 WETable 3. Temporal changes in vehicle fleets.Region (reference) No. of vehicles (% of total) 1998/1992Vehicle category Engine type 1992 1995 1998 RatioThuringiaaCars Total 751,673 1,165,813 1,204,304 1.60Gas 699,467 1,047,013 1,083,216 1.55Diesel 51,716 (7)b 83,695 (7) 96,278 (8) 1.86Buses + trucks Total 53,555 84,215 145,957 2.72Gas 9,222 12,901 20,176 2.19Diesel 44,236 (83) 80,786 (96) 125,315 (86) 2.83All categories Total 805,228 1,250,028 1,350,261 1.67Gas 708,689 1,059,914 1,103,392 1.56Diesel 95,952 (11) 164,481 (13) 221,593 (16) 2.31ErfurtcCars 80,553d 95,738 92,588 1.15eMotorcycles 17,644d 2,163 2,958 0.17eTrucks 2,381d 8,166 8,763 3.6eAll categories 104,012d 108,299 106,528 1.02eaData from Kraftfahrt-Bundesamt (30).bPercentages for Thuringia data refer to percentage of vehicle category composedof vehicles with diesel engines. cData from Statistisches Jahrbuch Theuringen 1996/97 (45). dNumber for 1990. eRatio for1998/1990. Figure 6. NO hourly profiles for 1991 and 1998, showing weekends (WE) andweekdays (WD) separately. 60.0050.0040.0030.0020.0010.000.00NO (µg/m3 )Hour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241991 WD1998 WD1991 WE1998 WEDiscussionTime–series plots revealed considerable sea-sonal variability of all pollutants in Erfurt,East Germany, with high concentrations dur-ing the winter months (October–March)compared to the summer months (April–September). Such seasonal patterns wereexpected considering the tendency for winterinversions in the Erfurt region and the use ofcoal and other fuels for heating during thewinter. Due to less complete combustionunder colder climatic conditions, automotiveemissions are usually also increased duringcolder months of the year. These data alsodemonstrate the drastic change in ambient airquality that has occurred in Erfurt after the1990 unification of East and West Germany.As direct methods of establishing andquantifying the sources contributing to par-ticulate matter, source apportionment tech-niques such as chemical mass balance orfactor analysis methods can be used. Lackinga complete emissions inventory or analysis ofparticle constituents, we were not able toperform a direct source analysis at this time.However, by using markers of sources forwhich we had data and analyzing how thesemarkers varied over the course of the day in1991 compared to 1998, our goal was toassess to the extent to which changes in spe-cific sources have affected various particlesize classes. Regressions between NC andSO2 and NO indicate major influences ofpower plants, residential coal combustion,and road traffic on the number concentra-tions of particles in Erfurt. A limitation inour analysis was the comparison of only twoseasons of data, although these where chosento be representative of a documented trendof changing air pollutant concentrations overthe 1990–1999 period.Before unification, lignite-fired powerplants in former East Germany operatedwith minimum emission controls (21). Coalconsumption data demonstrated that signifi-cant power plant retrofit programs andchanges in residential heating worked tosharply decrease SO2 concentrations inErfurt. Decreases in the use of coal werereflected in decreased levels of ambient SO2and PM2.5. Comparing to the current U.S.Environmental Protection Agency’s (U.S.EPA) National Ambient Air QualityStandards (NAAQS), PM2.5 levels were inexcess of the 65 µg/m3 standard 67 timesduring winter 1991 (October–March), com-pared to only 6 times during winter 1998.Similarly, the 24-hr NAAQS for SO2 of 365µg/m3 was exceeded 6 times during winter1991 (October–March) and was notexceeded at all after 1995. Our first hypothesis for the observedincreases in the smallest ultrafine particlesconsidered decreased ultrafine particle scav-enging due to fewer NC0.03–0.5 andNC0.5–2.5 particles in the ambient air. Dueto their high diffusivity, particles in theNC0.01–0.03 range are subject to enhancedaggregation particularly with larger particlesbecause of their larger cross-section. Thechange of the ratio of NC0.01–0.03 particles toPM2.5 (TSP) concentrations by a factor of 8(5), respectively, can be taken as a measureof such particle scavenging. Increases inNC0.01–0.03 per mass of TSP over time Articles • Changes in air quality and emissions in GermanyEnvironmental Health Perspectives • VOLUME 109 | NUMBER 4 | April 2001 331Figure 8. Standardized hourly profiles for particle number concentration andSO2 in 1991 (weekdays only). 200180160140120100806040200Percent of averaged medianHour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241991 NCO.O1–O.O31991 NCO.O3–O.O51991 NCO.5–2.51991 SO2 (µg/m3)Figure 9. Standardized hourly profiles for particle number concentration, NO,and CO in 1991 (weekdays only).Figure 10. Standardized hourly profiles for particle number concentration andSO2 in 1998 (weekdays only). 250200150100500Percent of averaged medianHour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241991 NCO.O1–O.O31991 NCO.O3–O.O51991 NCO.5–2.51991 NO (µg/m3)1991 CO (mg/m3)250200150100500Percent of averaged medianHour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241998 NCO.O1–O.O31998 NCO.O3–O.O51998 NCO.5–2.51998 SO2 (µg/m3)Figure 11. Standardized hourly profiles for particle number concentration, NO,and CO in 1998 (weekdays only). 250200150100500Percent of averaged medianHour1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 241998 NCO.O1–O.O31998 NCO.O3–O.O51998 NCO.5–2.51998 NO (µg/m3)1998 CO (mg/m3)Articles • Ebelt et al.332 VOLUME 109 | NUMBER 4 | April 2001 • Environmental Health Perspectivessuggests that decreased particle scavenging isa possible mechanism for retaining highnumbers of small particles in the air. Ouranalysis suggested that residential coal burn-ing was associated with afternoon peaks oflarger particles in 1991. This relationshipfaded over time due to decreases in coal con-sumption and consequent decreases in theNC0.03–0.5 and NC0.5–2.5 fractions. Ifdecreases in coal use have affected the con-centrations of large particles, then changes inthis emission source have potentially con-tributed to decreased scavenging of smallerparticles. The second hypothesis regarding theincrease in NC0.01–0.03 considered traffic asthe cause, and we focus the discussion of thishypothesis in the following paragraphs.Vehicle emissions were found to be associ-ated with the morning peaks of all particlesize fractions. However, significant differ-ences in the morning NC0.01–0.03 in particu-lar, between 1991 and 1998, suggested thatvehicle emissions in 1998 were directly pro-ducing larger numbers of NC0.01–0.03. Thisresult is supported by a study conducted inthe United Kingdom, which also found thatultrafine particles dominate the numbercount and that number count data give aclear indication of recent road traffic emis-sions (32). Supplemental data indicated that thetotal number of vehicles did not change inErfurt between 1991 and 1998; however,the composition of the vehicle fleet didchange. Although there are no available data,the proportion of two-stroke cars in theErfurt vehicle fleet has decreased dramati-cally since 1991. Additionally, trucksincreased in number by a factor greater than3, and motorcycles decreased by a factorgreater than 5. The percentage of vehicleswith diesel motors increased from 11% to16% between 1992 and 1998, doubling innumber. These changes are likely a combina-tion of increased diesel usage in all ofGermany and Western Europe as well as thesocietal and structural changes that havetaken place as a result of unification. Diesel emissions are a significant air pol-lution issue. Estimates from the U.S. EPAindicate that diesel vehicles contribute 27%of on-road NOx and > 60% of on-road PMemissions (33). Kirchstetter et al. (34) foundheavy-duty diesel trucks to be much higheremitters of NOx, PM2.5, SO42-, and blackcarbon than light-duty vehicles. Additionally,heavy-duty diesel engines were found to emit15–20 times more fine particles than light-duty vehicles per unit mass of fuel burned(34). There is also evidence suggesting dieselengines produce particles of smaller sizes.When assessing particles from all mobilesources, mass distributions peak between 0.1and 0.2 µm particle diameter (35). Of twomedium-duty diesel vehicles tested, Kleemanet al. (35) found particle number size distrib-utions peaking at 30 nm and 50 nm particlediameter. Several studies have demonstratednew diesel vehicles as higher emitters of ultra-fine particles in comparison to diesel engineswith older technology (32,36). For example,Bagley et al. (36) found that a 1998 dieselengine delivered 15–35 times the number ofparticles than a 1981 engine produced,although the total mass was reduced. Thiswas due to a 30- to 60-fold increase in thenumber of smaller primary particles (36).As stated in our second hypothesis,increases in the number of trucks and otherdiesel-powered vehicles may be contributingTable 4. General linear models predicting NC fractions (column headings indicate the dependent variables). Independent NC0.01–0.03 NC0.03–0.5 NC0.5–2.5 NC0.01–0.03 NC0.03–0.5 NC0.5–2.5variable t p t p t p Interaction t p t p t pIntercept 7.0654 0.0000 6.5667 0.0000 2.3433 0.0192 [Winter = 1991] × [hr = 2] –0.1878 0.8510 0.2027 0.8394 0.2192 0.8265[WINTER = 1991] –2.7877 0.0053 –0.1602 0.8727 0.4035 0.6866 [Winter = 1991] × [hr = 3] –0.1821 0.8555 –0.2520 0.8011 0.0289 0.9769[WINTER = 1998] [Winter = 1991] × [hr = 4] –0.0954 0.9240 –0.6444 0.5194 –0.0706 0.9437[WE_WD = 0] –13.6873 0.0000 –7.9740 0.0000 –4.8550 0.0000 [Winter = 1991] × [hr = 5] –0.4506 0.6523 –1.1000 0.2714 0.3192 0.7496[WE_WD = 1] [Winter = 1991] × [hr = 6] –1.7233 0.0849 –1.6520 0.0986 0.6005 0.5482[hr = 1] –0.5876 0.5569 –0.8024 0.4224 –0.1992 0.8422 [Winter = 1991] × [hr = 7] –3.6535 0.0003 –2.6818 0.0074 0.5908 0.5547[hr = 2] –0.4005 0.6888 –0.9770 0.3286 –0.1990 0.8422 [Winter = 1991] × [hr = 8] –2.5480 0.0109 –1.8715 0.0614 0.3387 0.7349[hr = 3] –0.0677 0.9460 –1.0935 0.2743 –0.1581 0.8744 [Winter = 1991] × [hr = 9] –3.5805 0.0003 –1.5650 0.1177 –0.6504 0.5154[hr = 4] –0.4670 0.6405 –1.0499 0.2939 –0.1826 0.8551 [Winter = 1991] × [hr = 10] –2.2048 0.0275 –1.0131 0.3111 –0.5403 0.5890[hr = 5] 0.5961 0.5512 –0.6479 0.5171 –0.4868 0.6264 [Winter = 1991] × [hr = 11] –1.5440 0.1227 –0.8481 0.3964 –1.7663 0.0774[hr = 6] 3.2035 0.0014 0.4958 0.6201 –0.2572 0.7970 [Winter = 1991] × [hr = 12] –1.2945 0.1956 0.6457 0.5185 –0.5659 0.5715[hr = 7] 6.8155 0.0000 2.4687 0.0136 –0.2490 0.8034 [Winter = 1991] × [hr = 13] –0.5716 0.5676 –0.3380 0.7354 –1.2032 0.2290[hr = 8] 5.9466 0.0000 3.8648 0.0001 –0.3009 0.7635 [Winter = 1991] × [hr = 14] –0.2602 0.7947 –0.3428 0.7317 –0.6994 0.4844[hr = 9] 7.5553 0.0000 3.4140 0.0006 –0.5550 0.5789 [Winter = 1991] × [hr = 15] –0.2225 0.8239 0.9604 0.3369 0.0107 0.9915[hr = 10] 5.4919 0.0000 3.1855 0.0015 –0.5865 0.5576 [Winter = 1991] × [hr = 16] –0.0235 0.9812 0.5460 0.5851 –0.2223 0.8241[hr = 11] 4.9918 0.0000 3.1058 0.0019 –0.3745 0.7081 [Winter = 1991] × [hr = 17] –0.3983 0.6904 0.5648 0.5722 –0.8899 0.3736[hr = 12] 5.0369 0.0000 2.3489 0.0189 –0.5219 0.6018 [Winter = 1991] × [hr = 18] –0.0210 0.9832 1.1870 0.2353 –1.6704 0.0949[hr = 13] 4.0586 0.0001 2.7164 0.0066 –0.0555 0.9558 [Winter = 1991] × [hr = 19] –0.7550 0.4503 1.5163 0.1295 –1.6625 0.0965[hr = 14] 3.8022 0.0001 2.5658 0.0103 –0.3450 0.7301 [Winter = 1991] × [hr = 20] –0.8757 0.3812 0.8772 0.3805 –0.6627 0.5076[hr = 15] 3.9226 0.0001 1.2383 0.2157 –0.8150 0.4152 [Winter = 1991] × [hr = 21] 0.0966 0.9230 0.5564 0.5780 –0.7743 0.4388[hr = 16] 3.5315 0.0004 1.2033 0.2289 –1.0870 0.2771 [Winter = 1991] × [hr = 22] –0.0650 0.9482 0.1081 0.9140 –0.6406 0.5218[hr = 17] 3.4745 0.0005 1.1518 0.2495 –0.9125 0.3616 [Winter = 1991] × [hr = 23] 0.7958 0.4262 0.6871 0.4921 0.0073 0.9942[hr = 18] 3.0559 0.0023 0.8676 0.3857 –0.8886 0.3743 [Winter = 1991] × [hr = 24][hr = 19] 3.6616 0.0003 0.9204 0.3574 –1.2229 0.2215 [Winter = 1998] × [hr = 1–24][hr = 20] 2.9145 0.0036 0.9693 0.3325 –1.3642 0.1726 [Winter = 1991] × [WE_WD = 0] 4.7869 0.0000 2.0255 0.0429 1.5249 0.1274[hr = 21] 0.9281 0.3534 –0.0476 0.9620 –1.4565 0.1454 [Winter = 1991] × [WE_WD = 1][hr = 22] 0.7690 0.4420 0.2630 0.7926 –1.1736 0.2406 [Winter = 1998] × [WE_WD = 0][hr = 23] –0.5833 0.5597 –0.4859 0.6271 –0.8348 0.4039 [Winter = 1998] × [WE_WD = 1][hr = 24] [Winter = 1991] × SO2 –9.6593 0.0000 –7.4273 0.0000 –10.1383 0.0000[Winter = 1998] × SO2SO2 9.1031 0.0000 9.6561 0.0000 15.5278 0.0000 [Winter = 1991] × NO –10.9911 0.0000 1.2034 0.2289 18.1812 0.0000[Winter = 1998] × NONO 34.7273 0.0000 57.7616 0.0000 7.3894 0.0000 [WE_WD = 0] × SO2 –1.3678 0.1715 –1.2127 0.2253 1.1589 0.2466[WE_WD = 1] × SO2[WE_WD = 0] × NO 1.6588 0.0973 3.0926 0.0020 1.6952 0.0901[WE_WD = 1] × NOAbbreviations: WE_WD = 0, weekend; WE_WD = 1, weekday. Independent variables, indicated in the rows, were regressed against each of the particle size fractions indicated in thecolumns. For the two periods, winter 1998 is compared to winter 1991 as the reference value. For day of the week, weekdays are compared to weekends as the reference value.directly to increased ultrafine particle concen-trations. Also, changes in engine technologyleading to combustion that is more completecould be a cause for increasing numbers ofultrafine particles. In a review of diesel emis-sions studies, Yanowitz et al. (37) reportedthat emissions of CO and PM have fallensteadily in the last 10 years which is consistentwith the findings discussed above (35,36).Engine technology that decreases PMincreases the efficiency of combustion,thereby also lowering CO and total hydrocar-bon emissions. However, no changes in aver-age NOx emissions of diesel vehicles havebeen observed (37,38). NOx and PM emis-sions are inversely correlated, which is themain barrier to lowering diesel emissions (37).In Erfurt, NO concentrations did notincrease along with the increase in the num-ber of trucks. In fact, NO concentrationsdecreased by > 50% between 1991 and1998. This was likely due to changes inengine technology that took place during the1990s. For example, introduction of thethree-way catalysts on gasoline-powered,light-duty vehicles has been mandatory onall new cars sold within the European Unionsince January 1993. A trend analysisbetween 1986 and 1994 for 15 Swedishcities confirmed that the real-world effi-ciency of the three-way catalyst correspondsto a reduction in NOx emissions from theaverage vehicle by at least 80–90% (39).Because the majority of NOx from automo-biles is in the form of NO, NO from thesevehicles has decreased over time. Therefore,the observed 50% decrease in NO, althoughless than expected from the introduction ofthree-way catalyst-equipped, light-duty vehi-cles, may be explained by the increasednumber of trucks with comparatively highNO emissions.Overall, decreases in the concentrationsof larger particle fractions were largely associ-ated with the decreases in coal combustionthat occurred during the 1990s. Changes invehicle composition and emissions werereflected in a changed hourly pattern ofNC0.01–0.03 and were thought to be the causeof increased number concentrations of theseultrafine particles. However, the exact contri-bution of vehicle and fuel types was not ana-lyzed, and emission rates have been found tovary significantly with vehicle classificationand driving conditions, among other parame-ters (35,37,40). There have been suggestionson how to distinguish between gasoline anddiesel vehicles in source apportionment stud-ies, for example, by using polycyclic aromatichydrocarbon profiles created to represent thelocal fleet of vehicles (41).REFERENCES AND NOTES1. Harris CD. Unification of Germany in 1990. Geogr Rev81:170–183 (1990). 2. Berentsen WH. Socioeconomic development of EastGermany, 1989-1998: a comparative perspective. Post-Soviet Geography and Economics 40(1):27–43 (1999).3. Commission of the European Communities. Integration ofthe East German economy into the community. Eur Econ46:187–198 (1990). 4. Acker K, Moeller D, Marquardt W, Brueggemann E,Wieprecht W, Auel R, Kalass D. Atmospheric researchprogram for studying changing emission patterns afterGerman unification. Atmos Environ 32:3435–3443 (1998).5. Umweltbundesamt und Statistisches Bundesamt, ed.Umweltdaten Deutschland 1998. Berlin:Umweltbundesamtund Statistisches Bundesamt, 1998. 6. Kraemer U, Behrendt H, Dogner R, Ranft U, Ring J, WillerH, Schlipkoeter H. Airway diseases and allergies in Eastand West German children during the first 5 years afterreunification: time trends and the impact of sulphur diox-ide and total suspended particles. Int J Epidemiol28:865–873 (1999).7. Heinrich J, Hoelscher B, Wichmann HE. Decline of ambi-ent air pollution and respiratory symptoms in children.Am J Respir Crit Care Med 161:1930–1936 (2000).8. Spix C, Heinrich J, Dockery D, Schwartz J, Völksch G,Schwinkowski K, Cöllen C, Wichmann HE. Air pollutionand daily mortality in Erfurt, East Germany, 1980–1989.Environ Health Perspect 101:518–526 (1993).9. Peters A, Goldstein IF, Beyer U, Franke K, Heinrich J,Dockery DW, Spengler JD, Wichmann HE. Acute healtheffects of exposure to high levels of air pollution in east-ern Europe. Am J Epidemiol 144:570–581 (1996).10. Wichmann H-E, Heinrich J. Health effects of high levelexposure to traditional pollutants in East Germany—review and ongoing research. Environ Health Perspect103(suppl 2):29–35 (1995).11. NRC. Research priorities for airborne particulate matter: I.Immediate priorities and a long-range research portfolio:Committee on Research Priorities for Airborne ParticulateMatter. Washington, DC:National Research Council, 1998.12. Seaton A, MacNee W, Donaldson K, Godden D.Particulate air pollution and acute health effects. Lancet345:176–178 (1995).13. Peters A, Wichmann HE, Tuch T, Heinrich J, Heyder J.Respiratory effects are associated with the number ofultrafine particles. Am J Respir Crit Care Med155:1376–1383 (1997).14. Peters A, Wichmann HE, Tuch T, Heinrich J, Heyder J.Comparison of the number of ultra-fine particles and themass of fine particles with respiratory symptoms in asth-matics. Ann Occup Hyg 41(Suppl 1):19–23 (1997).15. Wichmann HE, Peters A. Epidemiological evidence of theeffects of ultrafine particle exposure. Philos Trans R SocA 358:2751–2769 (2000). 16. Pekkanen J, Timonen KL, Ruuskanen J, Reponen A,Mirme A. Effects of ultrafine and fine particles in urbanair on peak expiratory flow among children with asth-matic symptoms. Environ Res 74:24–33 (1997).17. Tuch T, Brand P, Wichmann HE, Heyder J. Variation ofparticle number and mass concentration in various sizeranges of ambient aerosols in Eastern Germany. AtmosEnviron 31:4193–4197 (1997).18. Wichmann HE, Spix C, Tuch T, Wölke G, Peters A,Heinrich J, Kreyling W, Heyder J. Daily Mortality andFine and Ultrafine Particles in Erfurt, Germany. Part I:Role of Particle Number and Particle Mass. ResearchReport 98. Cambridge, MA:Health Effects Institute, 2000.19. Wichmann HE, Spix C, Wittmaack K, Cyrys J, Wölke G,Peters A, Heinrich J, Kreyling W, Heyder J. DailyMortality and Fine and Ultrafine Particles in Erfurt,Germany. Part II:  Role of Sources, ElementalComposition and Other Pollutants. Research Report.Cambridge, MA:Health Effects Institute, in press.20. Holman C. Sources of air pollution. In: Air Pollution andHealth (Holgate S, Samet J, Koren H, Maynard R, eds).London:Academic Press, 1999;115–148.21. Smith DJ. Germany emphasizes power plant pollutioncontrol. Power Eng 97:30–32 (1993).22. DOE/EIA. International Energy Outlook 1999 - TransportationEnergy Use. DOE/EIA - 0484 (99). Washington, DC:U.S.Department of Energy/Energy Information Administration,1999.23. Walsh MP. Global trends in motor vehicle pollution con-trol: 1996 state of the art and remaining problems. Int JVehicle Design 18:233–242 (1997). 24. Brauer M, Dumyahn TS, Spengler JD, Gutschmidt K,Heinrich J, Wichmann H-E. Measurement of acidic aerosolspecies in eastern Europe: implications for air pollutionepidemiology. Environ Health Perspect 103:482–488 (1995).25. Cyrys J, Gutschmidt K, Brauer M, Dumyahn T, Heinrich J,Spengler JD, Wichmann HE. Determination of acidic sulfateaerosols in urban atmospheres in Erfurt (F.R.G.) and Sokolov(former C.S.S.R.). Atmos Environ 29:3545–3557 (1995).26. Cyrys J, Heinrich J, Brauer M, Wichmann HE. Spatial vari-ability of acidic aerosols, sulfate and PM10 in Erfurt, EasternGermany. J Expo Anal Environ Epidemiol 8:447–464 (1998).27. Brand P, Gebhart J, Below M, Georgi B, Heyder J.Characterization of environmental aerosols on HelgolandIsland. Atmos Environ 25A:581–585 (1991).28. Tuch T, Mirme A, Tamm E, Heinrich J, Heyder J, Brand P,Roth C, Wichmann HE, Pekkanen J, Kreyling WG.Comparison of two particle size spectrometers for ambientaerosol measurements. Atmos Environ 34:146–169 (1999).29. Heywood JB. Internal Combustion Engine Fundamentals.New York:McGraw-Hill, 1988.30. Kraftfahrt-Bundesamt. Statistiche Mitteilungen desKraftfahrt-Bundesamtes Reihe 1. Heft 11/1999. Flensburg,Germany:Das Kraftfahrt-Bundesampt, 1999.31. Buzorius G, Hameri K, Pekkanen J, Kulmala M. Spatialvariation of aerosol number concentration in Helsinkicity. Atmos Environ 33:553–565 (1999).32. Harrison RM, Jones M, Collins G. Measurements of thephysical properties of particles in the urban atmosphere.Atmos Environ 33:309–321 (1999).33. U.S. EPA. National Air Pollutant Emission Trends Update,1970–1997. Washington, DC:U.S. EnvironmentalProtection Agency, 1998.34. Kirchstetter TW, Harley RA, Kreisberg NM, StolzenburgMR, Hering SV. On-road measurement of fine particleand nitrogen oxide emissions from light- and heavy-dutymotor vehicles. Atmos Environ 33:2955–2968 (1999).35. Kleeman MJ, Schauer JJ, Cass GR. Size and composi-tion distribution of fine particulate matter emitted frommotor vehicles. Environ Sci Technol 34:1132–1142 (2000).36. Bagley ST, Baumgard KJ, Gratz LD, Johnson JJ, LeddyDG. Characterization of Fuel and Aftertreatment DeviceEffects on Diesel Emissions. Cambridge, MA:HealthEffects Institute, 1996.37. Yanowitz J, McCormick RL, Graboski MS. In-use emis-sions from heavy-duty diesel vehicles. Environ SciTechnol 34:729–740 (2000).38. Cadle SH, Gorse RAJ, Bailey BK, Lawson DR. Real-worldvehicle emissions: a summary of the ninth coordinatingresearch council on-road vehicle emissions workshop. JAir Waste Manag Assoc 50:278–291 (2000).39. Sjodin A, Sjoberg K, Svanberg PA, Backstrom H.Verification of expected trends in urban traffic NOx emis-sions from long-term measurements of ambient NO2 con-centrations in urban air. Sci Total Environ 189/190:213–220(1996).40. Shi JP, Mark D, Harrison RM. Characterization of parti-cles from a current technology heavy-duty diesel engine.Environ Sci Technol 34:748–755 (2000).41. Cadle SH, Mulawa PA, Hunsanger EC, Nelson K, RagazziRA, Barrett R, Gallagher GL, Lawson DR, Knapp KT,Snow R. Composition of light-duty motor vehicle exhaustparticulate matter in the Denver, Colorado area. EnvironSci Technol 33:2328–2339 (1999).42. Schmidt M. Personal communication. 43. Schubert E. Personal communication. 44. Strom und Fernwaerme GmbH. Unpublished data. 45. Thuringer Landesanstalt fuer Umwelt, ed. StatistischesJahrbuch Thueringen 1996/97. Erfurt, Germany:ThuringerLandesanstalt fuer Umwelt, 1997.Articles • Changes in air quality and emissions in GermanyEnvironmental Health Perspectives • VOLUME 109 | NUMBER 4 | April 2001 333


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items