Time Horizon DependentCharacterization Factors forAcidification in Life-CycleAssessment Based on Forest PlantSpecies Occurrence in EuropeR O S A L I E V A N Z E L M , * , †

M A R K A . J . H U I J B R E G T S , † , ‡

H A N S A . V A N J A A R S V E L D , §

G E R T J A N R E I N D S , | | D I C K D E Z W A R T , #

J A A P S T R U I J S , ‡ , # A N DD I K V A N D E M E E N T † , #

Department of Environmental Sciences, Institute for Waterand Wetland Research, Radboud University Nijmegen,P.O. Box 9010, 6500 GL, Nijmegen, The Netherlands, LCAexpertise centre, National Institute of Public Health and theEnvironment, P.O. Box 1, 3720 BA, Bilthoven,The Netherlands, The Netherlands EnvironmentalAssessment Agency (MNP), P.O. Box 303, 3720 AH Bilthoven,The Netherlands, AlterrasGreen World Research, P.O. Box 47,6700 AA, Wageningen, The Netherlands, Laboratory forEcological Risk Assessment, National Institute of PublicHealth and the Environment, P.O. Box 1,3720 BA, Bilthoven, The Netherlands

This paper describes a new approach in life-cycle impactassessment to derive characterization factors foracidification in European forests. Time horizon dependentcharacterization factors for acidification were calculated,whereas before only steady-state factors were available.The characterization factors indicate the change in thepotential occurrence of plant species due to a change inemission, and they consist of a fate and an effectfactor. The fate factor combines the results of anatmospheric deposition model and a dynamic soil acidificationmodel. The change in base saturation in soil due to anatmospheric emission change was derived for 20, 50, 100,and 500 year time horizons. The effect factor was basedon a dose-response curve of the potential occurrence ofplant species, derived from multiple regression equationsper plant species. The results showed that characterizationfactors for acidification increase up to a factor of 13from a 20 years to a 500 years time horizon. Characterizationfactors for ammonia are 4.0-4.3 times greater thanthose for nitrogen oxides (NOx), and characterizationfactors for sulfur dioxide are 1.4-2.0 times greater thanthose for NOx. Aggregation of damage due to acidificationwith other impact categories on the European scalebecomes feasible with the applied approach.

IntroductionLife-Cycle Assessment is a tool that deals with the environ-mental impacts associated with a product or service (1). Itreviews the complete life cycle in different steps. An importantpart of this process is the life cycle impact assessment (LCIA),where inventory data are converted into impact indicatorsfor various impact categories. Udo de Haes et al. (2) give anoverview of impact categories, which comprises, amongothers, acidification. Atmospheric deposition of inorganicsubstances, such as sulfates and nitrates, cause a change inacidity in the soil. In the process from acid deposition on soilto ecosystem damage, two major sources of response delayare distinguished. Biogeochemical processes can delaychemical response to acid deposition in soil. Depending onthe chemical status of a soil, atmospheric input of acidifyingsubstances is neutralized by a number of buffer reactions(3). Biological processes can delay the response of indicatororganisms, such as damage to trees. Consequently, it ispossible that acidity keeps changing in the ecosystem, whiledeposition is already constant or even decreasing (4). Foralmost all plant species, there is a clearly defined optimumof acidity. A serious deviation in acidity from this optimumleads to a lower occurrence probability for that specificspecies and is referred to as acidification. As a result, changesin levels of acidity will cause shifts in species occurrence (5,6). Important acidifying emissions are nitrogen oxides (NOx),ammonia (NH3), and sulfur dioxide (SO2) (2, 6).

A commonly used impact indicator in LCIA is thecharacterization factor. Characterization factors for acidi-fication express the contribution of each emission into theenvironment to acidification of the ecosystem, and they areobtained at either midpoint or endpoint level (7). Midpointsare defined as points anywhere on the cause-effect chainbetween emission release and acidification, e.g., changes insoil acidity. Endpoints are defined as points indicating theconsequences of the effect: damage to the ecosystem causedby acidification, e.g., loss of biodiversity.

Characterization factors for acidification are commonlyaddressed on a midpoint level (5, 8-16). A number of authorscalculated country-specific characterization factors for acidi-fication in Europe on the basis of the critical load concept(9-12), while Norris retrieved region-specific characterizationfactors for Northern America and Mexico on the basis oftotal terrestrial deposition (14). The endpoint level providesinformation that directly matters to society, and at this level,consequences of the different impact categories can easilybe compared and added up in terms of ecosystem quality(5). Hayashi et al. (6) focused on the development of a functionthat relates emission of acidifying substances to endpointdamage for Japan. Effects on net primary production ofvegetation were quantified. Goedkoop and Spriensma (5)applied the Dutch Nature Planner (17) that focuses on thepercentage of threatened species in The Netherlands causedby acidifying emissions. Steen (18) retrieved global charac-terization factors on midpoint level (based on base cationcapacity) as well as endpoint level (based on extinction ofspecies).

The aim of this paper is to calculate characterizationfactors for acidification on an endpoint level. Calculationswere based on 240 plant species in forest ecosystems on aEuropean scale. The focus was on forests, since European-wide information on plant species composition of the groundvegetation was available for this type of vegetation only. Thisstudy combines the strength of the current midpoint models,which have a relatively large spatial coverage, with the

* Corresponding author phone: 0031-24-3652060; fax: 0031-24-3653030; e-mail: r.vanzelm@science.ru.nl.

† Radboud University Nijmegen.‡ LCA expertise centre, National Institute of Public Health and the

Environment.§ The Netherlands Environmental Assessment Agency.| AlterrasGreen World Research.# Laboratory for Ecological Risk Assessment, National Institute of

Public Health and the Environment.

Environ. Sci. Technol. 2007, 41, 922-927

922 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007 10.1021/es061433q CCC: $37.00 2007 American Chemical SocietyPublished on Web 01/05/2007

endpoint models, which describe actual damage of acidifyingsubstances. The characterization factors for acidification wereaddressed in terms of a marginal change in potentially notoccurring fraction of plant species (PNOF) in European forestsdue to a marginal change in emission of respectively NOx,NH3, and SO2 to air. Here, the PNOF is defined as the fractionof plant species that does not occur in a European forestregion because of acidifying conditions. To show the ap-plicability of the characterization factors, normalizationscores for acidification in European forests were alsopresented. Furthermore, as acidification is a process thatcan occur over a long time scale (>100 years), characterizationfactors and normalization scores were specified for a numberof time horizons.

Materials and MethodsCalculation Procedure. Characterization Factors for acidi-fication of substance x (CFx in m2‚yr‚kg-1) are defined as themarginal change in the potentially not occurring fraction ofplant species in European forest area j (dPNOFj) due to amarginal change in emission of acidifying substance x (dMx

in kg‚yr-1) in Europe. The size of forest area j (Aj in m2) isapplied as a weighing factor:

The process of converting emissions to air into ecosystemdamage caused by acidification can be subdivided in a fateanalysis step, linking marginal changes in emissions tomarginal changes in base saturation (BS), and an effectanalysis step, linking marginal changes in BS to marginalchanges in increased potentially not occurring fraction ofplant species.

where DEPj is the deposition of acid equivalents in forestarea j (eq‚ha-1‚yr-1), and BSj is the degree to which theadsorption complex of a soil in forest area j is saturated withbase cations, being Ca, Mg, and K (-). When more base cationsare present, the buffer capacity of soil toward acidifyingsubstances is higher. Changes in BS in mineral soil caninfluence the occurrence of plant species in forests (19).

The fate factor was calculated in two steps. First, changesin acid deposition, derived from changes in air emission,were calculated. Second, changes in BS, derived from changesin acid deposition, were calculated. Characterization factorsfor acidification were calculated for time horizons of 20, 50,100, and 500 years.

Atmospheric Fate Factor. The atmospheric part of thefate factor (FFatm,j in eq‚ha-1‚kg-1) was defined as follows:

where Tx,Europefj is the transfer coefficient of acidifyingsubstance x from source area Europe to forest area j(eq‚ha-1‚kg-1). Europe was divided in 8064 receptor areas ofabout 50 × 50 km, each area characterized by its uniquecoordinates, land use class, roughness length, and forest area.The atmospheric fate model EUTREND (20, 21) was used tocalculate depositions for each receptor area caused byEuropean emissions of acidifying substances. EUTRENDcovers the entire European continent with its marginal seasand can be used to model transport and deposition on the

local, regional, and continental scale. EUTREND combinesa Gaussian plume model, describing short-range, localtransport, and dispersion, with a Lagrangian trajectory model,describing long-range transport (22). The model has beenvalidated on several scales (20, 23, 24). Emission, dispersion,advection, chemical conversion, and dry and wet depositionare accounted for in the model. Meteorological data of 1990,a meteorological moderate year, were applied. Tx,Europefj wasobtained as a source-receptor vector for, respectively, NOx,NH3, and SO2, containing dDEPj/dMx for each receptor area.

Soil Fate Factor. The soil part of the fate factor (FFsoil inha‚yr‚eq-1) was defined as the marginal change in BS due toa marginal change in deposition in forest area j:

The soil fate factor depends on multiple parameters, e.g.,deposition of the acidifying substance, hydrology, andbiogeochemistry, in a potentially nonlinear way. The deposi-tion profiles derived with EUTREND were used as input forthe calculation of the soil fate factor. To obtain a yearlybackground scenario, emissions were obtained from 1990through 2000 and emission estimates from 2001 through 2010(25-27). For 2011 through 2500, constant emissions wereassumed, equal to the estimated ones from 2010. Appliedemissions are listed in the Supporting Information. Thesimulation model for acidification’s regional trends, version2 (SMART2, (28-30)) was used to numerically determine∆BSj/∆DEPj. Emission increments of 1, 5, and 10% from thereference year 2000 were used to check for linearity in thisrange. The model can calculate soil fate factors on a yearlybasis. SMART2 is a single-layer soil acidification and nutrientcycling model for forest soils. It is based on the model SMART(31) and extended with a nutrient cycle (litter fall, miner-alization, and uptake) and an improved modeling of hydrol-ogy, including runoff and upward and downward solutefluxes. SMART2 captures the major hydrological and bio-geochemical processes in vegetation, litter, and mineral soil.The results of SMART2 were validated in previous studies(32-34).

Before application of SMART2, it was calibrated for eachreceptor by iteratively running the model from 1880 through1995 until BS in the topsoil in 1995 was correctly simulated.Actual values of BS in 1995 were obtained for approximately5000 forest observation points in Europe. SMART2 wasapplied to about 130 000 different forest-soil combinationsfor Europe, constructed by overlaying maps that includeinformation about forest and soil types, historical aciddeposition, climate data estimates, and climate zones.

Effect Factor. The dimensionless effect factor of forestarea j (EFj) was defined as follows:

The potentially not occurring fraction of plant species wasderived from the probability of occurrence of an individualplant species s (Ps). Multiple regression can be used to expressthe occurrence probability of individual species as a functionof variability in predefined environmental factors and possiblytheir interactions (35-37). Here, multiple regression equa-tions were used as developed by De Vries et al. (19):

where x1 to xn are environmental explanatory variables, such

CFx ) ∑j

(Aj‚dPNOFj

dMx) (1)

FFatm,j )dDEPj

dMx) Tx,Europefj (3)

FFsoil,j )dBSj

dDEPj≈ ∆BSj

∆DEPj(4)

EFj )dPNOFj

dBSj(5)

ln( Ps

1 - Ps) ) a + b1x1 + c1x1

2 + b2x2 + c2x22 + .... +

bnxn + cnxn2 (6)

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 923

as stand and site characteristics, climatic and depositionvariables, and soil chemical data, and a...cn are regressionconstants.

Equation 6 is simplified to relate only base saturation toPs:

where as reflects the actual situation of all environmentalexplanatory variables except BS relevant for plant species sat a specific forest area, and bs and cs are species-specificregression constants related to BS.

To express variability among forest areas, a Monte Carlosimulation was executed and 1000 as values were derived,representing 1000 hypothetical forest areas. De Vries et al.(19) derived κ-values for a number of plant species, such thatwhen Ps > κ, a plant species is present in a specific forestarea, and when Ps < κ this plant species does not occur. Theκ-value was derived by finding, per species, the optimalcorrespondence between the multiple regression equationand the empirical occurrence of a species (0 or 1) over alllocations included in the dataset. Per species, an imaginaryκ was varied between 0 and 1, to convert the continuousprediction in occurrence/ no occurrence for all locations.Per species, the imaginary κ that gave the best overallcorrespondence with the empirical dataset was set as thespecies-specific κ. For 240 plant species, we then checked,from their κ-values, whether they occur at each of the 1000simulated forest areas and at different BS levels of the soil.κ, as, bs, and cs values for the 240 plant species are given inthe Supporting Information. Every individual plant speciesshows its own dose-response curve of its occurrenceprobability. These individual dose-response behaviors arenow combined, resulting in the potentially occurring fractionof plant species at a specific BS-level (POFBS):

where No,BS is the number of simulated species-forest areacombinations when a species occurs at a specific BS, and Nt

is the total number of species-forest area combinations,which equals 240 000 in our study.

To calculate the effect factor, the PNOF is derived by thefollowing:

It is possible that the occurrence of a specific plant speciesin a forest area is not dependent on BS, but that the naturalbackground situation (without acidifying components) isalready unsuited for this species. To account for thisbackground situation the potentially not occurring fractionof plant species due to added acidifying emissions (PNOFadded)was calculated by subtracting the potentially not occurringfraction that is not related to BS, PNOFbackground, from PNOFBS,and scaled to species occurring at the background situation(38):

Figure 1 shows the dose-response relationship of the addedfraction of species not occurring in European forests as afunction of BS of the soil. The obtained best fit follows thefollowing equation:

with an explained variance (R2) of 1.00. It can be seen thatthe linear function holds for BS larger than ( 0.15.

ResultsThe derivative of eq 11, dPNOFadded/dBS, represents the effectfactor for acidification in European forest soils. For envi-ronmental BS-fractions larger than 0.15, the effect factorequals -0.26. The acidification effect factor for BS < 0.15 is0, implying that there are no changes in the diversity of specieswhen base saturation is varied within this range.

As the acidification effect factor for BS appears to belocation-independent for BS > 0.15, a location-independentfate factor for acidification (m2‚yr‚kg-1) can be calculated:

Fate factors are calculated for BS > 0.15 to stay in line withthe effect factor. Fate and characterization factors foracidification of 1% emission increase are shown in Table 1.Since they differ to a maximum of a factor of 1.05 from theemission scenarios 5 and 10%, no distinction is made for thedifferent emission increases. Fate and characterization factorsafter 500 years are larger than fate factors after 20 years byfactors of 9, 13, and 13 for SO2, NOx, and NH3, respectively.Fate and characterization factors for NH3 are 4.0-4.3 timeslarger than for NOx, and factors for SO2 are 1.4-2.0 timeslarger than for NOx.

Normalization scores for acidification in European forestswere derived by multiplication of the characterization factorswith the substance-specific acidifying emissions in Europeof year 2000 (see Supporting Information). Results aredisplayed in Figure 2 and show the potentially not occurring

ln( Ps

1 - Ps) ) as + bs × BS + cs × BS2 (7)

POFBS )No,BS

Nt(8)

PNOFBS ) 1 - POFBS (9)

PNOFadded )PNOFBS - PNOFbackground

1 - PNOFbackground(10)

PNOFadded ) 0.27 - 0.26 × BS (11)

FIGURE 1. A dose response function of the potentially not occurringfraction of plant species due to acidifying emissions (PNOFadded) asa function of base saturation (BS) in mineral soil. The fitted linearfunction follows PNOFadded ) 0.27 - 0.26 × BS with an explainedvariance R2 ) 1.00 and holds for BS larger than 0.15.

TABLE 1. Effect, Fate, and Characterization Factors forAcidificationa

effectfactor (-)

fate factor(m2‚yr/kg)

characterization factor(m2‚yr/kg)

timehorizon 20 50 100 500 20 50 100 500

NOx -0.26 -0.11 -0.24 -0.46 -1.41 0.03 0.06 0.12 0.37NH3 -0.26 -0.43 -1.04 -1.98 -5.73 0.11 0.27 0.52 1.49SO2 -0.26 -0.22 -0.47 -0.81 -1.98 0.06 0.12 0.21 0.51

a The characterization factors express the disappeared fraction ofplant species per unit emission of an acidifying substance multipliedwith the actual forest area in Europe.

FFx )

∑j

(∆BSj × Aj)

∆Mx

(12)

924 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007

fraction of plant species over a certain area due to emissionsin year 2000 of NOx, NH3, and SO2 respectively (km2).Normalization scores are given in the Supporting Informa-tion. It can be seen that, over 500 years time horizon fromthe three acidifying substances, SO2 is the main cause ofacidification of forest soils in Europe.

DiscussionCharacterization Factors. In this study characterizationfactors for acidification are obtained on an endpoint level,indicating marginal change of damage to European forestecosystems caused by acidification. Characterization factorsfor acidification were calculated for various time horizons.The main advantage of the endpoint approach is that it makesmeaningful aggregation of the consequences of differentimpacts possible. An extensive comparison of different levelsof impact and the advantage of combining them is given byBare and Gloria (39). Characterization factors for acidificationon the endpoint level are applicable as reference points fordecision makers, as they reveal information on the actualdamage to ecosystems. The estimation of European-basedcharacterization factors can, for example, be applied whenthe damage due to newly emitted acidifying substances inEurope is of interest without knowing the exact sourcelocation beforehand. The normalization scores show thepotentially not occurring fraction of plant species over acertain area due to acidifying emissions of NOx, NH3, or SO2

in year 2000. This opens up the possibility of identifying theimportance of acidification compared to other impactcategories, such as ecotoxicity (40, 41) and land-use (42).

Comparison to Other Studies. Variation in characteriza-tion factors for acidification between studies is caused bydifferences in framework, chosen impact category indicator,input parameters, and modeling approach. The calculatedcharacterization factors for acidification after 500 yearsapproach the steady-state situation and are compared tosteady-state factors that were derived in studies of Goedkoopand Spriensma (5), Hayashi et al. (6), Steen (18), Potting (11),Hettelingh et al. (8), Huijbregts et al. (9), and Seppala et al.(12). To be able to compare the characterization factors foracidification, the contributions of NOx, NH3, and SO2 to theimpact score if 1 kg of each acidifying substance is emitted,are compared (Figure 3).

The methods of Potting (11), Hettelingh et al. (8),Huijbregts et al. (9), and Seppala et al. (12) show rathercomparable shares of NOx, NH3, and SO2 to the impact scoreas our research. Shares of NOx are between 13 and 16%,shares of NH3 are between 44 and 63%, and shares of SO2 arebetween 22 and 40%. The results from Goedkoop andSpriensma (5), Hayashi et al. (6), and Steen (18), however,do not show the same pattern. The deviation by Goedkoopand Spriensma (5) may be explained by the fact that part of

the fate of acidifying substances was neglected; transportprocesses and weather conditions were not taken intoaccount. Hayashi et al. (6) explain that their characterizationfactor for acidification by NH3 is less representative thantheir factors obtained for NOx and SO2 because of largeruncertainties due to limited data availability to calculate theatmospheric fate factor. The characterization factors derivedby Steen (18) are world-average, in contradiction to theEuropean or country-specific factors, derived with the othermethods.

Model Characteristics and Findings. In contrast to thecurrently used steady-state models, we applied a dynamicmodel. Biogeochemical and biological processes can causea delay of decades or even centuries before steady-state isreached. In this context, dynamic models are useful becausethey attempt to estimate the time evolution of soil responsesto changes in acid deposition and can be used to assess thetime required for a new (steady-) state to be achieved (4).Policy makers can use dynamic model results to set out policyon the short term as well as on the long term.

NH3, NOx, and SO2 are the acidifying pollutants includedin the study. Acidifying emissions of HF, HCl, H2S, and H3-PO4 hardly play a role on the continental scale and aretherefore excluded (8). It must be kept in mind, however,that on regional and local scale, emissions of these substancescan play a larger role, which is, for example, the case for HFin Japan (6).

Emissions in many life cycles partly occur outside Europe.It should be noted that these emissions cannot be charac-terized in a valid way with characterization factors foracidification estimated with our European model. Emissionand deposition can also differ among and even withincountries (11). In case life cycle emissions are situated in afew regions, site-dependent characterization factors foracidification are preferred in LCIA (11). Since the location ofemissions and forest areas are known and used as input inthe fate models, it would, in principle, be possible to calculateregion-specific characterization factors for acidification. Forthis purpose, substance-specific source-receptor matriceson a country level for the atmospheric transport modelEUTREND need to be developed and combined with theSMART2 model.

Information was available for plant species in forests.heathlands, meadows, and other terrestrial ecosystem typescan be included in this type of analysis as well, provided thatresponse and fate data are available. Information wascurrently available for 240 forest plant species, neglectingother species relevant in forest areas. A general effect factorfor all species was calculated. It is also possible to focus ontarget species, which are considered in need of protection,in the compilation of a dose response curve.

FIGURE 2. Normalization scores (km2) for acidification, representingthe potentially not occurring fraction of plant species over a certainarea due to emissions in year 2000 of NOx, NH3, and SO2 respectively.

FIGURE 3. Contributions of NOx, NH3, and SO2 to the impact scorewhen 1 kg of each acidifying substance is emitted, as derived fromour study, Goedkoop and Spriensma (5), Hayashi et al. (6), Steen(18), Potting (11), Hettelingh et al. (8), Huijbregts et al. (9), and Seppa1la1et al. (12).

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 925

Fate factors were calculated for the total BS range as wellto check whether significant differences are observed fromthe calculations for the range BS > 0.15. Fate factor outcomesare 1% higher for NOx and NH3 emissions, and 2-3% higherfor SO2 emissions, which are considered to be relatively smalldistinctions. Thus, the forest area in Europe where BS is below0.15 appears to have a low influence on the fate factors andcharacterization factors compared to the area where BS islarger than 0.15.

In forest soils with higher pH levels (pH > ca. 5.5),deposition of acidifying substances hardly influences basesaturation (31), because the incoming acidity is buffered byprotonation of variable charges or incongruent weatheringof the parent material. Deposition of acidifying substancesmay result in a pH-change, which is not followed by adecrease in base saturation. The pH change due to increasein acid deposition for pH > 5.5 may influence speciesoccurrence, which is not covered by the current modelapproach. However, it should be noted that a pH above 5.5is rare for forest soils in Europe. Measurements at 6000 forestplots in Europe show that more than 80% of the forests havea pH below 5.5 (43). Our model calculations showed that95% of the forest areas in Europe had a pH below 5.5 in year2000. Furthermore, SMART2 does not include increaseddiffusion of calcium from deep sources after input ofacidifying substances in acid soils emanated from calcium-rich parent material (44). These soils occur only on a limitedarea within Europe. Including this environmental mechanismmay result in slightly lower fate factors and, thereby,characterization factors compared to the current results.

SO2 will reach steady-state before NOx and NH3, becausesulfur immediately leads to acidification when deposited anddoes not interact with vegetation. Change in nitrogendeposition leads to changes in nitrogen contents in thevegetation. It takes decades until a new equilibrium in theecosystem is reached, which causes a delay in reaching asteady-state level in the soil (45). Nitrogen contents in thesoil keep changing for a longer period of time, contributingto a larger increase in characterization factor for acidificationcompared to SO2, from 20 to 500 years time horizon (seeTable 1).

In LCIA, interest generally lies on characterization factorswhich can be applied to small emission changes, a so-calledmarginal approach. Hettelingh et al. (8) discussed theappropriateness of the -10 and +10% increments used bythemselves, Potting (11), and Krewitt et al. (10). They statethat these methods were not designed to evaluate smallchanges in emissions and are, therefore, not appropriate tocalculate characterization factors. The spatial resolution ofthese models, especially with regard to emission, atmospherictransport, and deposition, is not high enough to allow suchanalyses. Our study included emission increases between 1and 10%. This paper reveals that characterization factors foracidification, based on PNOF, hardly change when emissionsincrements change. Seppala et al. (12) also came to the sameconclusion for emission changes between 0.5 and 50%.

We computed characterization factors for acidificationon an endpoint level. Characterization factors for acidificationincrease over time, due to the decreasing buffer capacity ofthe soil. Characterization factors for NH3 are 4.0-4.3 timeslarger than for NOx and factors for SO2 are 1.4-2.0 timeslarger than for NOx. Our normalization scores show that,based on emissions of year 2000, SO2 contributes most toacidification over 500 years time-horizon. Our procedurecombines the strength of the current midpoint models thathave a relatively large spatial coverage, with the endpointmodels that describe actual damage of acidifying substances.A meaningful aggregation of damage in Europe caused bydifferent impact categories is now possible.

AcknowledgmentsThis work is part of the ReCiPe project under the name ofthe Dutch LCA Expertise Centre and was financed by theMinistry of Housing, Spatial Planning, and Environment.

Supporting Information AvailableInformation on yearly emissions, plant species, and nor-malization scores. This material is available free of chargevia the Internet at http://pubs.acs.org.

Literature Cited(1) Consoli, F.; Allen, D.; Boustead, I.; Fava, J.; Franklin, W.; Jensen,

A. A.; De Oude, N.; Parrish, R.; Perriman, R.; Postlethwaite, D.;Quay, B.; Sequin, J.; Vigon, B. Guidelines for Life-Cycle Assess-ment: A “Code of Practice”; Society of Environmental Toxicologyand Chemistry: Pensacola, Fl, 1993.

(2) Udo de Haes, H. A.; Finnveden, G.; Goedkoop, M. J.; Hauschild,M.; Hertwich, E. G.; Hofstetter, P.; Jolliet, O.; Klopfer, W.; Krewitt,W.; Lindeijer, E.; Muller-Wenk, R.; Olsen, S. I.; Pennington, D.W.; Potting, J.; Steen, B. Life-Cycle Impact Assessment: StrivingTowards Best Practice; SETAC: Pensacola, Fl, 2002.

(3) Blaser, P.; Zysset, M.; Zimmermann, S.; Luster, J. Soil acidificationin Southern Switzerland between 1987 and 1997: A case studybased on the critical load concept. Environ. Sci. Technol. 1999,33, 2383-2389.

(4) De Vries, W.; Reinds, G. J.; Posch, M.; Sanz, M. J.; Krause, G. H.M.; Calatayud, V.; Renaud, J. P.; Dupouey, J. L.; Sterba, H.; Vel,E. M.; Dobbertin, M.; Gunderson, P.; Voogd, J. C. H. IntensiveMonitoring of Forest Ecosystems in Europe, 2003, TechnicalReport; EC/ UN/ECE: Brussels, Geneva, 2003.

(5) Goedkoop, M. J.; Spriensma, R. The Eco-Indicator ’99: A Damage-Oriented Method for Life-Cycle Impact Assessment, 1999/36A;Ministry of Housing, Spatial Planning, and Environment: TheHague, The Netherlands, 1999.

(6) Hayashi, K.; Okazaki, M.; Itsubo, N.; Inaba, A. Development ofdamage function of acidification for terrestrial ecosystems basedon the effect of aluminum toxicity on net primary production.Int. J. LCA 2004, 9, 13-22.

(7) Jolliet, O.; Muller-Wenk, R.; Bare, J. C.; Brent, A.; Goedkoop, M.J.; Heijungs, R.; Itsubo, N.; Pena, C.; Pennington, D. W.; Potting,J.; Rebitzer, G.; Stewart, M.; Udo de Haes, H. A.; Weidema, B.The LCIA midpoint-damage framework of the UNEP/SETAClife cycle initiative. Int. J. LCA 2004, 9, 394-404.

(8) Hettelingh, J. P.; Posch, M.; Potting, J. Country-dependentcharacterization factors for acidification in Europe; a criticalevaluation. Int. J. LCA 2005, 10, 177-183.

(9) Huijbregts, M. A. J.; Schopp, W.; Verkuijlen, E.; Heijungs, R.;Reijnders, L. Spatially explicit characterization of acidifying andeutrophying air pollution in life-cycle assessment. J. Ind. Ecol.2000, 4, 75-92.

(10) Krewitt, W.; Trukenmuller, A.; Bachmann, T. M.; Heck, T.Country-specific damage factors for air pollutants. Int. J. LCA2001, 6, 199-210.

(11) Potting, J. Spatial Differentiation in Life Cycle Assessment, Ph.D.thesis. University of Utrecht, Utrecht, 2000.

(12) Seppala, J.; Posch, M.; Johansson, M.; Hettelingh, J. P. Country-dependent characterisation factors for acidification and ter-restrial eutrophication based on accumulated exceedance asan impact category indicator. Int. J. LCA 2006, 11, 403-416.

(13) Bare, J. C.; Norris, G. A.; Pennington, D. W.; McKone, T. TRACI;the tool for the reduction and assessment of chemical and otherenvironmental impacts. J. Ind. Ecol. 2003, 6, 49-78.

(14) Norris, G. A. Impact characterization in the tool for the reductionand assessment of chemical and other environmental impacts;methods for acidification, eutrophication and ozone formation.J. Ind. Ecol. 2003, 6, 79-101.

(15) Heijungs, R.; Guinee, J. B.; Huppes, G.; Lankreijer, R. M.; Udode Haes, H. A.; Wegener Sleeswijk, A.; Ansems, A. M. M.; Eggels,A. M. M.; Van Duin, R.; De Goede, H. P. Environmental LifeCycle Assessment of Products. Guidelines and Backgrounds;Centre of Environmental Sciences: Leiden, The Netherlands,1992.

(16) Steen, B. A Systematic Approach to Environmental PriorityStrategies in Product Development (EPS). Version 2000sGeneralSystem Characteristics; Chalmers University of Technology:Goteborg, Sweden, 1999.

926 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 3, 2007

(17) Latour, J. B.; Staritsky, I. G.; Alkemade, J. R. M.; Wiertz, J. DeNatuurplanner, Decision Support Systeem Natuur en Milieu,Versie 1.1, RIVM 711901019; RIVM: Bilthoven, The Netherlands,1997.

(18) Steen, B. A Systematic Approach to Environmental PriorityStrategies in Product Development (EPS), Version 2000sModelsand Data of the Default Method; Chalmers University ofTechnology: Goteborg, Sweden, 1999.

(19) De Vries, W.; Reinds, G. J.; Van Dobben, H.; De Zwart, D.; Aamlid,D.; Neville, P.; Posch, M.; Auee, J.; Voogd, J. C. H.; Velet, E. M.Intensive Monitoring of Forest Ecosystems in Europe. TechnicalReport 2002; EC, UN/ECE: Brussels, Geneva, 2002.

(20) Van Jaarsveld, J. A. Modelling the Long-Term AtmosphericBehaviour of Pollutants on Various Spatial Scales, Ph.D. thesis.University of Utrecht, Utrecht, 1995.

(21) Van Jaarsveld, J. A.; Van, Pul, W. A. J.; De, Leeuw, F. A. A. M.Modelling transport and deposition of persistent organicpollutants in the European region. Atmos. Environ. 1997, 7,1011-1024.

(22) Bleeker, A.; Draaijer, G. P. J.; Klap, J. M.; Van Jaarsveld, J. A.Deposition of Acidifying Components and Base Cations inGermany in the Period 1987-1995, RIVM 7221080027; RIVM:Bilthoven (The Netherlands), 2000.

(23) Asman, W. A. H.; Van Jaarsveld, H. A. A variable-resolutiontransport model applied for NHx in Europe. Atmos. Environ.1992, 26A, 445-464.

(24) Baart, A. C.; Diederen, H. S. M. A. Calculation of the AtmosphericDeposition of 29 Contaminants to the Rhine Catchment Area,TNO R 91/219; TNO: Delft, The Netherlands, 1991.

(25) EU. Richtlijn 2001/81/EG van het Europees Parlement en de Raadvan 23 oktober 2001 inzake nationale emissieplafonds voorbepaalde luchtverontreinigende stoffen; European Union, 2001.

(26) UN/ECE. Protocol to the 1979 Convention on Long-RangeTransboundary Air Pollution to Abate Acidification, Eutrophi-cation and Ground-Level Ozone; UN/ECE: Geneva, 2000.

(27) Vestreng, V. EMEP/MSC-W Technical Report, Review and Revi-sion, Emission data reported to CLRTAP. MSC-W Status Report2003, EMEP/MSC-W Note 1/2003, ISSN 0804-2446; EMEP/MSCW-W: Blindern, Norway, 2003.

(28) Kros, J. Evaluation of Biogeochemical Models at Local andRegional Scale. Ph.D. thesis. Wageningen University, Wagenin-gen (The Netherlands), 2002.

(29) Van Hinsberg, A.; Kros, H. Dynamic modelling and the calcula-tion of critical loads for biodiversity. In: Posch, M., De Smet,P. A. M., Hettelingh, J.-P., Downing, R. J.; Calculation andMapping of Critical Thresholds in Europe. Status Report 2001;Coordination Centre for Effects, RIVM: Bilthoven, The Neth-erlands, 2001; pp 73-80.

(30) Kros, J.; Reinds, G. J.; De Vries, W.; Latour, J. B.; Bollen, M.Modelling the response of terrestrial ecosystems to acidificationand desiccation scenarios. Wat. Air Soil Pollut. 1995, 85, 1101-1106.

(31) De Vries, W.; Posch, M.; Kamari, J. Simulation of the long-termsoil response to acid deposition in various buffer ranges. WaterAir Soil Pollut. 1989, 48, 349-390.

(32) Ahonen, J.; Rankinen, K.; Holmberg, M.; Syri, S.; Forsius, M.Application of the SMART2 model to a forested catchment in

Finland: comparison to the SMART model and effects ofemission reduction scenarios. Boreal Environ. Res. 1998, 3, 221-233.

(33) Wamelink, G. W. W.; Van Oene, H.; Mol-Dijkstra, J. P.; Kros, J.;Van Dobben, H. F.; Berendse, F. Validatie van de modellenSMART2, SUMO 1, NUCOM en MOVE op site-, regionaal ennationaal niveau, Alterra 065; Wageningen: The Netherlands,2001.

(34) Mol-Dijkstra, J. P.; Kros, H. Modelling effects of acid depositionand climate change on soil and run-off chemistry at Risdalsheia,Norway. Hydrol. Earth Syst. Sci. 2001, 5, 487-498.

(35) Bakkenes, M.; De Zwart, D.; Alkemade, J. R. M. MOVE nationaalModel voor de Vegetatie versie 3.2. Achtergronden en analysevanmodelvariantenRIVM408657006(inDutch);RIVM: Bilthoven,The Netherlands, 2002.

(36) Ertsen, A. C. D.; Frens, J. W.; Nieuwenhuis, J. W.; Wassen, M.J. An approach to modelling the relationship between plantspecies and site conditions in terrestrial ecosystems. LandscapeUrban Planner 1995, 31, 143-151.

(37) Latour, J. B.; Reiling, R. A multiple stress model for vegetation(‘move’): a tool for scenario studies and standard-setting. Sci.Total Environ. 1993, Supplement (2), 1513-1526.

(38) Klepper, O.; Bakker, J.; Traas, T. P.; Van de Meent, D. Mappingthe potentially affected fraction (PAF) of species as a basis forcomparison of ecotoxicological risks between substances andregions. J. Hazard. Mater. 1998, 61, 337-344.

(39) Bare, J. C.; Gloria, T. P. Critical analysis of the mathematicalrelationships and comprehensiveness of life cycle impactassessment approaches. Environ. Sci. Technol. 2006, 40, 1104-1113.

(40) Huijbregts, M. A. J.; Struijs, J.; Goedkoop, M.; Heijungs, R.;Hendriks, A. J.; Van de Meent, D. Human population intakefractions and environmental fate factors of toxic pollutants inlife cycle impact assessment. Chemosphere 2005, 61, 1495-1504.

(41) Van de Meent, D.; Huijbregts, M. A. J. Calculating life-cycleassessment efect actors from potentially affected fraction-basedecotoxicological response functions. Environ. Toxicol. Chem.2005, 24, 1573-1578.

(42) Kollner, T. Species-pool effect potentials (SPEP) as a yardstickto evaluate land-use impacts on biodiversity. J. Clean. Prod.2000, 8, 293-311.

(43) Vanmechelen, L.; Groenemans, R.; Van Ranst, E. Forest SoilCondition in Europe. Results of a Large-Scale Soil Survey; EC-UN/ECE: Brussels, Geneva, 1997.

(44) Grigall, D. F.; Ohmann, P. R. Calcium and forest systems:diffusion from deep sources. Soil Sci. 2005, 170, 129-136.

(45) Posch, M.; Hettelingh, J. P.; Slootweg, J. Manual of DynamicModelling of Soil Response to Atmospheric Deposition, RIVM259101012; RVIM: Bilthoven, 2003.

Received for review June 15, 2006. Revised manuscript re-ceived November 14, 2006. Accepted November 16, 2006.

ES061433Q

VOL. 41, NO. 3, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 927