2003), Modeling terrestrial biogenic sources of oxygenated organic emissions, Earth Interac

Copyright ?2003.Paper 7-007;35,091words,5Figures,2Tables..

https://www.sodocs.net/doc/bf13564350.html,

Modeling Terrestrial Biogenic

Sources of Oxygenated Organic Emissions

Christopher Potter

NASA Ames Research Center,Moffett Field,California

Steven Klooster

California State University,Monterey Bay,Seaside,California

David Bubenheim and Hanwant B.Singh

NASA Ames Research Center,Moffett Field,California

Ranga Myneni

Boston University,Boston,Massachusetts

Received 11February 2003;accepted 17April 2003

ABSTRACT:In recent years,oxygenated volatile organic chemicals

(OVOCs)like acetone have been recognized as important atmospheric con-

stituents due to their ability to sequester reactive nitrogen in the form perox-

yacetyl nitrate (PAN)and to be a source of hydroxyl radicals (HO x )in critical regions of the atmosphere.The potential biogenic sources of acetone include

terrestrial plant canopies,oxidation of dead plant material,harvest of culti-

vated plants,biomass burning,and the oceans.These sources are poorly con-

*Corresponding author address:C.Potter,NASA Ames Research Center,

Moffett Field,CA

E-mail address:cpotter@https://www.sodocs.net/doc/bf13564350.html,

strained at present in budgets of atmospheric chemistry.Based on reported

laboratory,?eld,and satellite observations to date,an approach is presented

for a biosphere model to estimate monthly emissions of acetone from the

terrestrial surface to the atmosphere.The approach is driven by observed land

surface climate and estimates of vegetation leaf area index(LAI),which are

generated at0.5o spatial resolution from the NOAA satellite Advanced Very

High Resolution Radiometer(AVHRR).Seasonal changes in LAI are esti-

mated using the Moderate Resolution Imaging Spectroradiometer(MODIS)

radiative transfer algorithms to identify the probable times and locations of

crop harvest in cultivated areas and leaf fall of newly dead plant material in

noncultivated areas.Temperature-dependent emission factors are applied to

derive global budgets of acetone?uxes from terrestrial plant canopies,oxi-

dation of dead plant material,and harvest of cropland plants.The predicted

global distribution of acetone emissions from live foliage is strongly weighted

toward the moist tropical zones,where relatively warm temperatures and high

LAI are observed in rain forest areas year-round.Predicted acetone emissions

are estimated at between54and172Tg yr?1from live foliage sources and

between7and22Tg yr?1from decay of dead foliage.These?ux totals from

vegetation are large enough to account for the majority of postulated biogenic

acetone sources in models of global atmospheric chemistry,but our model

predictions are subject to veri?cation in subsequent?ux control experiments

using a variety of plant species,particularly those from humid tropical zones.

KEYWORDS:Acetone,Biosphere,Model

1.Introduction

Oxygenated volatile organic chemicals(OVOCs)have been found to be ubiqui-tous and abundant components of the global troposphere(Singh et al.,2001). Among the myriad of such chemicals present,acetone and methanol are the most dominant.Acetone has been recognized as an important compound in atmospheric chemistry because it can sequester reactive nitrogen in the form of peroxyacetyl

nitrate(PAN,CH

3CO

2

ONO

2

)and provide a direct source of hydroxyl radicals

HO

x (OH+HO

2

)in the upper troposphere(Singh et al.,1995).These acetone-

initiated reactions can substantially alter the rates of ozone formation in the upper troposphere(Wennberg et al.,1998;Collins et al.,1999;Muller and Brasseur, 1999),and therefore play a role in the Earth’s radiation balance.The sources and sinks of OVOCs are poorly known but biogenic emissions are thought to provide an important source.

Recent assessments have projected that between300and500Tg of OVOCs are emitted into the atmosphere every year(Guenther et al.,1995;Fall,1999). Large fractions of emissions are suspected to be biogenic in origin,including about30–40Tg yr?1of acetone directly from terrestrial biogenic sources,ex-cluding biomass burning and oxidation of organic precursors(Singh et al.,2000; Singh et al.2001;Jacob et al.,2002).Development of accurate,seasonally re-solved emission inventories,ideally based on high-resolution ecological mapping studies,is crucial to the improvement of global atmospheric model simulations and the consistency between these model predictions and atmospheric measure-

ments of acetone concentrations.However,no global gridded acetone or methanol emissions inventories have yet been provided through cooperative international research programs,such as the Global Emissions Inventory Activity(GEIA)com-ponent of the International Global Atmospheric Chemistry(IGAC)Project,a core project of the International Geosphere–Biosphere Program,or through the Emis-sion Database for Global Atmospheric Research(EDGAR;Olivier et al.,2001).

There have been only a few published measurements for biogenic emissions of acetone and methanol.Previously reported rates of acetone emission(from apparently undamaged plant tissues)range from0.01to0.5?g g?1h?1,whereas emission rates for methanol can exceed10?g g?1h?1(MacDonald and Fall,

1993).Land cover and land-use change may be particularly important factors to include for accurate estimation of OVOC emissions on regional scales.Cutting and dehydration of plant material in crop and timber harvest areas,lawn mowing, and range land grazing are some of the major human activities that are thought to impact global OVOC emissions(Fall,1999;de Gouw et al.,1999).

The aim of this study is to develop an approach for the?rst fully integrated global model to predict monthly biosphere–atmosphere?uxes of OVOC,with particular emphasis on major terrestrial vegetation sources of acetone.

Our speci?c objectives are to address the following research questions.

1.How do regional OVOC emissions vary with temperature,moisture avail-

ability,and vegetation properties?

2.How do regional OVOC emissions vary with plant tissue damage or plant

harvest by cutting?

3.How can predicted OVOC emissions be scaled up to regional and global

levels using remote sensing(satellite)data to drive ecosystem carbon models?

2.Modeling approach

2.1.Acetone emission algorithms

Here we describe the concept and algorithm development for a generalized model of biogenic OVOC emissions directly from terrestrial ecosystems.The model considers emissions from intact(living)plant canopies,decaying plant material, and harvested plant material from agricultural ecosystems.We have not consid-ered emissions from the oxidation of organic precursors.

For the most part,OVOC emission measurements in plant and ecosystem ?eld studies to date suggest that emission rates are controlled primarily by tem-perature effects on leaf growth and respiration metabolism(Fall,1999).Relative humidity and radiation?ux could be secondary controllers of these emissions, but their importance is poorly understood at present.Consequently,the OVOC emission algorithms we describe here are developed primarily as functions of basal emission rates(measured at303K),adjusted by the density of foliar bio-mass,a vegetation-speci?c emission coef?cient,and an exponential temperature response.Based chie?y on measurements reported by Schade and Goldstein (Schade and Goldstein,2001)and deGouw et al.(deGouw et al.,1999),we de?ne terrestrial OVOC source emissions by the following generalized algorithms.

Figure https://www.sodocs.net/doc/bf13564350.html,parison of OVOC emission equations for acetone as functions of

temperature,based on parameter de?nitions from Schade and Gold-

stein (Schade and Goldstein,2001)

Intact plant canopies (live foliage),

F ??D exp[?(T ?T )].

(1)

lf lf a ref Decaying plant material (dead foliage),

F ??D exp[?(T ?T )]P .

(2)df df a ref Harvested plant material (harvest foliage),

F ??D exp[?(T ?T )].(3)

ht ht a ref Here F emission ?uxes (in mg m ?2h ?1)are adjusted to monthly ?ux totals;?is the OVOC-speci?c base emission rate (in mg m ?2h ?1at 303K);D is the scaling factor for live foliar density,dead organic matter density,or harvested foliar density [in units of leaf area index (LAI),m 2m ?2];?is the vegetation-speci?c emission coef?cient;T a is surface air temperature (K);T ref ?303K;and P is the coef?cient for rainfall wetting effects on emissions from decaying plant material (P ?1,if rainfall ?1cm in a month;P ?2,if rainfall ?1cm in a month).

A comparison of the three OVOC emission equations as functions of tem-perature is shown in Figure 1.For acetone and methanol emissions,Schade and

Goldstein(Schade and Goldstein,2001)reported?values for live foliage that were several times larger than?values for dead foliage(Table1).This results in a more rapid exponential increase in emission rates with increasing temperature in the live foliage response function than in the dead foliage response function. Lacking more speci?c information,Equation(3)for harvest foliage emission uses the same?values as Equation(1)for live foliage.

Table1.Parameter values for OVOC emissions,as de?ned in Equations(1)–(3).All reported values of?base emission rates have been normalized to LAI?1,and are reported in units of mg C m?2h?1at303K.The vegetation-speci?c emission coef?cient is denoted by?.Pinus ponderosa values are from Schade and Gold-stein(Schade and Goldstein,2001).Trifolium ripens values are from Gouw et al. (Gouw et al.,1999).Mixed herbs values are from Fukui and Doskey(Fukui and Doskey,1998).NA is not available.

Live foliage Dead foliage Harvested foliage Species OVOC??????Pinus ponderosa Acetone0.1760.1100.0320.020

Pinus ponderosa Methanol 1.3670.1100.0320.030

Pinus ponderosa Ethanol0.3050.140NA NA

Trifolium ripens Acetone0.00608NA Trifolium ripens Methanol0.00152NA Mixed herbs Acetone0.047

Mixed herbs Methanol0.157

In all cases,we have normalized the reported base emission rate values of ?to LAI?1.The D scaling factor can then be applied to multiply emissions up to full vegetation canopy levels.For example,LAI for ponderosa pine has been reported by Schade and Goldstein(Schade and Goldstein,2001),which has been used as a scaling factor for other grid cells(using remotely sensed LAI).

The algorithm for OVOC emissions from decaying plant material[Equation (2)]is based on the assumption that emission from dead foliage derives mainly from organic material that has fallen onto the soil surface within the past year. These emissions would involve relatively rapid microbial decomposition of the most labile foliar compounds,leaving behind mainly structural leaf components such as cellulose and lignin that can remain in the soil organic horizons for years afterward in a relatively inert state.If this assumption is correct,then we can estimate the density and monthly timing of dead foliage transferred from live vegetation to the soil surface within the past year as a function of the maximum LAI value measured over the course of the year,multiplied by the monthly percent of annual leaf fall total.To compute the monthly D scaling factor in Equation (2),we use the formulation from Potter et al.(Potter et al.,1993)to compute the monthly percent of leaf fall based on a normalized seasonal pro?le of LAI values. Because this formulation for predicting the monthly percent of leaf fall is based on satellite remote sensing of global vegetation dynamics(see section2.2),we have the potential to capture many effects of large-scale natural disturbance events such as major wind storms,and frost,hail,and ice damage on the transfer of live to dead foliage pools.

Although corroborating evidence on moisture effects is not well developed,

Warneke et al.(Warneke et al.,1999)reported that acetone emission rates from

dead plant material were doubled when rainfall was recorded in a preceding24-

h period.We have included this presumed effect of wetting on rates of microbial

decay of dead plant material and OVOC emissions using the monthly precipitation

amount as an index of moisture delivered to the layers of dead plant material

decaying on the soil surface.Studies of rainfall interception in mature forests

generally indicate that vegetation canopies have a maximum storage capacity of

about0.2cm of rainfall(Federer,1979).For soil surface layers to remain moist

throughout a monthly time interval,multiple rainfall events greater than0.2cm

are necessary to offset evaporative losses.Therefore,a threshold value of rainfall ?1cm month?1was set in Equation(2)as a general index for wet season(versus dry season)moisture effects on decay rates of dead plant material on the soil

surface.

Emissions from harvested foliage are treated separately from naturally de-

caying foliage because laboratory measurements suggest that intentional damage

to previously intact live foliage can result in a transient but signi?cant increase

in OVOC emission rates(Kirstine et al.,1998).This potential crop harvest source

for acetone therefore results from anthropogenic activities that lead to the wound-

ing of cultivated plant foliage at the end of a crop growing season.On the basis

of measurements reported by deGouw et al.(deGouw et al.,1999),it is assumed

that harvest emissions have a maximum duration of7.5h.This maximum duration

time of elevated emissions following crop harvest is close to or slightly longer

than duration times for elevated emissions of a variety of VOC compounds re-

ported in other leaf wounding experiments(Fall et al.,1999).We note that the

harvest foliage source[Equation(3)]for OVOC does not include the subsequent

decay source from crop residues in the months following harvest.

2.2.Satellite and climate input datasets

Global climate drivers for our OVOC emission model include monthly mean

surface temperature and precipitation amount.These model inputs are long-term

(1931–60)averages gridded at0.5?spatial resolution(Legates and Willmott,

1990).The climatology was developed using a database of17,986independent

terrestrial station records and6955oceanic gridpoint records.Weather station data

were interpolated to a0.5?latitude–longitude grid using a spherically based in-

terpolation algorithm.

Global vegetation cover types in the model were determined by DeFries et

al.(DeFries et al.,1998)on the basis of differences in seasonal pro?les of the

Normalized Difference Vegetation Index(NDVI)gridded at0.5?resolution.NDVI

is a unitless parameter(scaled from0to1000)computed from the ratio of visible

(VIS)and near-infrared(NIR)sensor channels.Global coverage of NDVI is ob-

tained from the Advanced Very High Resolution Radiometer(AVHRR)satellite

sensor as this VIS:NIR channel ratio.As a terrestrial‘‘greenness’’index,the

AVHRR NDVI has been closely correlated with a variety of vegetation parame-

ters,including canopy LAI(Running and Nemani,1988;Sellers et al.,1994;

DeFries et al.,1995).

For derivation of LAI at0.5?spatial resolution,nearly complete AVHRR

datasets for the1980s and1990s have been produced from the National Atmo-spheric and Oceanic Administration(NOAA)Global Area Coverage(GAC)Level 1B data.These data consist of re?ectances and brightness temperatures derived from the?ve-channel cross-track scanning AVHRR aboard the NOAA Polar Or-biter‘‘afternoon’’satellites(NOAA-7,-9,-11,and-14).Monthly composite da-tasets remove much of the contamination due to cloud cover present in the daily AVHRR data(Holben,1986).Nevertheless,additional processing of the satellite imagery is necessary to eliminate remaining artifacts(Los et al.,1994).These data show minimal correlations with equatorial crossing times of the NOAA sat-ellites(Malmstro¨m et al.,1997),which suggests that corrections have been made for orbital drifts and switches between satellites(e.g.,NOAA-9to NOAA-11).

The input of live foliar density values to our model[Equation(1)]is made in terms of LAI derived from modi?ed Moderate Resolution Imaging Spectro-radiometer(MODIS)radiative transfer algorithms(Buermann et al.,2001),using the Global Invenory Monitoring and Modelling System(GIMMS)AVHRR data from1982to1999.The global MODIS algorithm products we use include im-proved calibration for intra-and intersensor variations,partial atmospheric cor-rection for gaseous absorption and scattering,and correction for stratospheric aerosol effects associated with volcanic eruptions.The three-dimensional transport equation based on an atmospherically corrected bidirectional re?ectance distri-bution function(BRDF)is used to simulate canopy re?ectances using sun-angle geometry and canopy/soil patterns as fundamental inputs.These advances are signi?cant because(i)NDVI relations are sensitive to changes in sun angle,view angle,and background re?ectance,while the MODIS algorithm actually exploits these changes to retrieve LAI;and(ii)the NDVI-based algorithm can only use two spectral bands,while the MODIS algorithm can ingest all the available spec-tral information to improve the quality of retrievals(Knyazikhin et al.,1998).

The monthly fraction of canopy litterfall and amounts of dead foliage trans-ferred to the soil surface for decay[for Equation(2)]were computed using the formulation of Potter et al.(Potter et al.,1993)using seasonal pro?les of MODIS algorithm LAI at0.5?spatial resolution from global average AVHRR data,as reported by Buermann et al.(Buermann et al.,2001).The month of crop harvest at0.5?spatial resolution was determined from the maximum monthly percent of leaf fall computed from the seasonal LAI-based formulation of Potter et al.(Potter et al.,1993)for those areas of the land surface classi?ed as croplands by DeFries et al.(DeFries et al.,1998).This cropland cover class delineated by DeFries et al.(DeFries et al.,1998)does not include tree harvest areas or expansive range lands that are grazed periodically by livestock.Therefore,we could not yet in-clude the effects of timber cutting,‘‘slash and burn’’deforestation,nor livestock damage to foliage in the equation[Equation(3)]for harvest-related emissions of OVOC.

3.Global model results

Peak predicted emissions of acetone from live foliage occur during the period of July–August from a variety of forested ecosystems(Figure2).Temperate decid-uous forests exhibit the most pronounced seasonal cycle in predicted acetone

Figure2.Predicted monthly?uxes of acetone from live foliage at four selected forest sites:boreal coniferous(Manitoba,Canada);temperate conifer-

ous(Thompson,OR,United States);temperate broadleaf deciduous

(Coweeta,NC,United States);tropical broadleaf evergreen(Manaus,

Brazil).

emissions,owing to the combined effect of warm summers/cold winters and leaf shedding in the fall.Tropical evergreen forests are predicted to respond most notably to seasonal changes in leaf area coverage rather than tempera-ture.

Predicted global emission of acetone from live foliage[Equation(1)]range from54to172Tg yr?1,depending on the selected setting for?,the base emission rate(Table2).Because there are not enough measurement datasets that include ?values to make a separate vegetation-speci?c assignment for each class in the land cover system of DeFries et al.(DeFries et al.,1998),our annual acetone emission totals are based on the assumption of globally uniform?values,from the lower estimate of54Tg yr?1using uniform?values for‘‘mixed herb’’veg-etation,to the higher estimate of172Tg yr?1using uniform?values for conif-erous forest vegetation(Table1).In either case of a uniform?setting,the global distribution of acetone emissions from live foliage is strongly weighted toward the moist tropical zones between23?N and23?S latitude(Figure3a).This zonal pattern derives from relatively warm temperatures and high LAI observed in tropical rain forest areas year-round.

Predicted acetone emissions from dead foliage were estimated at22Tg yr?1

(Table2).The global distribution of acetone emissions from dead foliage is close

to being evenly weighted between the areas of northern latitude zones(above 45?N)and moist tropical zones(Figure3b).Predicted acetone emissions totals from dead foliage are in?uenced by wide seasonal variations in temperature in the northern latitudes,and by the relative supply and autumn timing of dead leaf material delivered to the soil surface.

Table2.Predicted acetone emissions from global land cover types.Live foliage emissions are reported for two settings of the base emission rate,?(Table1):‘‘Low’’from Fukui and Doskey(Fukui and Doskey,1998)and‘‘High’’from Schade and Goldstein(Schade and Goldstein,2001).Dead foliage emissions are reported for two possible settings of P as de?ned in Equation(2).

Live foliage Dead foliage

Acetone emission Acetone emission

(Tg yr?1)(Tg yr?1)

Land cover class Land area

(DeFries et al.,1998)(km2?106)Low High P?1or2P?1 Broadleaf evergreen forest1425.882.6 6.1 1.9 Coniferous evergreen forest13 2.78.6 3.0 1.0 High-latitude woodland 5.70.4 1.30.70.2 Tundra70.10.40.30.1 Mixed deciduous–evergreen forest 6.6 2.1 6.7 1.90.6 Wooded grassland and savanna2212.840.9 4.6 1.8 Temperate grassland21 2.27.1 1.30.5 Bare ground170.10.20.00.0 Cropland13 3.812.0 2.40.8 Broadleaf deciduous forest3 3.310.7 1.50.5 Shrubs and bare ground110.6 1.80.20.1 Total133******** Owing to uncertainties in reported moisture controls on emissions from dead foliage,we tested the implications of overestimating the available moisture control on OVOC emissions by a setting P?2for all months when monthly precipitation amount is?https://www.sodocs.net/doc/bf13564350.html,ing instead a globally uniform setting of P?1in Equation (2),the predicted acetone source from dead foliage dropped to7Tg yr?1.We ?nd that these results are fairly sensitive to the precipitation?1cm month?1 threshold,because most productive areas of the land receive precipitation in ex-cess of this threshold.

Predicted acetone emissions from cropland harvest sources were relatively small at16.9?106g yr?1.We?nd that this predicted OVOC contribution is not large enough on a global basis to signi?cantly augment predicted acetone emis-sions from dead foliage in noncropland ecosystems,especially in the temperate climate zones.We hypothesize that neither the global extent of cropland harvest areas nor the relatively short duration(several hours)of elevated OVOC emis-sions following leaf wounding is large enough to make up important contri-butions to the combined biosphere sources of acetone emissions to the at-mosphere.This hypothesis may be proven false if most OVOC emissions from live and dead foliage sources are found not to be active year-round,24h day?1.

To demonstrate the versatility of this modeling approach and report an ad-ditional prediction for OVOC emissions from live foliage sources,we repeated

Figure3.Predicted acetone emissions(in g C m?2yr?1)from(a)live foliage and

(b)dead foliage.

the global model computations for methanol emissions with Equation(1),using the?base emission parameter value reported from Fukui and Doskey(Fukui and Doskey,1998)and the?emission parameter value reported from Schade and Goldstein(Schade and Goldstein,2001;Table1).Global biogenic sources from live foliage for methanol are predicted at254Tg yr?1using this model.Our predicted annual emission source from the terrestrial biosphere approaches earlier predictions from Guenther et al.(Guenther et al.,1995),who suggested the pri-mary biogenic source is more than280Tg yr?1of methanol.However,Holzinger et al.(Holzinger et al.,2000)have reported a laboratory study with Mediterranean holm oak that shows about one-tenth the methanol emissions reported by others for alfalfa,soybeans,pines/conifers,and aspen,suggesting the need for further

measurements of?and?parameter values.

Figure3.(Continued)

4.Discussion

The lower range value for our predicted emission total from live foliage(54Tg yr?1)accounts for the majority of previously postulated biogenic acetone sources in models of global atmospheric chemistry(Brasseur et al.,1998;Jacob et al., 2002).Singh et al.(Singh et al.,2000)estimated total global acetone emissions of56Tg yr?1.In a global atmospheric study of acetone sources,Jacob et al. (Jacob et al.,2002)assumed a direct emission from live vegetation to be20–40 Tg yr?1,and an additional emission source of9Tg yr–1from decaying plant residues worldwide.These estimates for dead foliage sources are close to our own predictions,but generally below the lower ends of our predicted?ux ranges for acetone for both live and dead foliage sources.One possible explanation for this

discrepancy is that Jacob et al.(Jacob et al.,2002)did not consider the high level

Figure4.Predicted acetone emissions from live foliage(dark line,right axis)and dead foliage(light line,left axis)(in Tg C yr?1)for0.5latitude zones.

of heteorogenity in foliar density observed by satellites across global ecosystems from tundra to grasslands to forests,nor from high latitudes to the moist tropical zones.Tropical evergreen forests have high LAI canopy levels of4–6year-round, which are exposed to fairly constant warm temperatures that may be conducive to elevated OVOC emission levels not routinely measured previously in(mostly temperate zone)ecosystems and plant species.

Another possible explanation for the high range in our predicted emission total of acetone from live foliage sources is the assumption of high and globally uniform?values,derived from a small number of reported?ux experiments in vegetation communities that are not representative of other plant species around the world.To expand the database of measured values,we are currently making new closed-chamber estimates of all parameters used in Equations(1)–(3)for acetone emissions from a variety of plant types,including broadleaf temperate and tropical tree species.In due time,representative measured values of?and?for all land cover types listed in Table2will replace uniform settings for these terms in global model algorithms.

Other potential improvements in our global model for biogenic OVOC emis-

sions will include con?rmation and inclusion of measured effects of the following:

1.moisture stress and variable humidity levels on emissions from live fo-

liage;

2.elevated acetone emissions from live foliage during the periods of bud

break(i.e.,triglyceride mobilization;Fall,1999)or during leaf senes-cence;and

3.wetting and drying on emissions from decaying foliage.

It is worth noting that the actual escape ef?ciency of any VOC to the above-canopy air mass is not likely to be100%(Guenther,1999),which means that the leaf-to-canopy-level emission?uxes estimated in the model algorithms presented here account only for emission?ux prior to(re)entrainment of OVOC by vege-tation canopies.Subsequent?eld and airborne measurements are necessary to determine if the initial biogenic?ux estimates we report from this modeling study are substantially diminished in magnitude prior to actual release into the free troposphere.

We conclude that predicted biogenic?ux totals from our modeling approach are large enough to account for the majority of postulated acetone sources from land in global atmospheric budgets.Although more measurement data for bio-genic OVOC?uxes will improve the vegetation-speci?c parameters in subsequent versions of the model,new global satellite observations of canopy leaf dynamics, which form the basis for our emissions approach,represent progress toward cap-turing important sources of variability in biogenic emissions over seasons and years.

Acknowledgments.This work was supported by the NASA Earth Observing System In-terdisciplinary Science Program,Grant MDAR-0044–0126,to Christopher Potter,principal in-vestigator.

References

Brasseur,G.P.,D.A.Hauglustaine,S.Walters,P.J.Rasch,J.-F.Muller,C.Granier,and X.X.

Tie,1998:MOZART,a global chemical transport model for ozone and related chemical tracers.1—Model description.J.Geophys.Res.,103,28,265–28,289.

Buermann,W.,J.Dong,X.Zeng,R.B.Myneni,and R.E.Dickinson,2001:Evaluation of the utility of satellite-based vegetation leaf area index data for climate simulations.J.Clim.,14, 3536–3550.

Collins,W.J.,D.S.Stevenson,C.E.Johnson,and R.G.Derwent,1999:Role of convection in determining the budget of odd hydrogen in the upper troposphere.J.Geophys.Res.,104, 26,927–26,941.

DeFries,R.,et al.,1995:Mapping the land surface for global atmosphere–biosphere models: Toward continuous distributions of vegetation’s functional properties.J.Geophys.Res.,100, 20,867–20,882.

DeFries,R.,M.Hansen,J.R.G.Townshend,and R.Sohlberg,1998:Global land cover classi-?cations at8km spatial resolution:the use of training data derived from Landsat imagery in decision tree classi?ers.Intl.J.Remote Sens.,19,3141–3168.

deGouw,J.A.,C.Howard,T.Custer,and R.Fall,1999:Emissions of volatile organic compounds from cut grass and clover are enhanced during the drying process.Geophys.Res.Lett.,26, 811–814.

Fall,R.,1999:Biogenic emissions of VOCs from higher plants.Reactive Hydrocarbons in the Atmosphere,edited by C.N.Hewitt,Academic Press,San Diego,43–96.

Fall,R.,T.Karl,A.Hansel,A.Jordan,and W.Lindinger,1999:Volatile organic compounds after leaf wounding:On-line analysis by proton-transfer-reaction mass spectrometry.J.Geophys.

Res.,104,15,963–15,974.

Federer,C.A.,1979:A soil–plant–atmosphere model for transpiration and availability of soil water.Water Resour.Res.,15,555–562.

Fukui,Y.,and P.V.Doskey,1998:Air-surface exchange of nonmethane organic compounds at a grassland site:Seasonal variations and stressed emissions.J.Geophys.Res.,103,13,153–13,168.

Guenther,A.,1999:Modeling biogenic volatile organic compound emissions to the atmosphere.

Reactive Hydrocarbons in the Atmosphere,edited by C.N.Hewitt,Academic Press,San Diego,41–94.

Guenther,A.,et al.,1995:A global model of volatile organic compound emissions.J.Geophys.

Res.,100,8873–8892.

Holben,B.N.,1986:Characteristics of maximum-value composite images from temporal AVHRR data.Intl.J.Remote Sens.,7,1417–1434.

Holzinger,R.,A.Jordan,A.Hansel,and W.Lindinger,2001:Methanol measurements in the lower troposphere near Innsbruck(47?16’N;11?24’E),Austria.Atmos.Environ.,35,2525–2532. Jacob,D.J.,B.D.Field,E.M.Jin,I.Bey,Q.Li,J.A.Logan,R.M.Yantosca,and H.B.Singh, 2002:Atmospheric budget of acetone.J.Geophys.Res.,107(D10)4100,doi:10.1029/ 2001JD000694.

Kirstine,W.,I.Galbally,Y.Ye,and M.Hooper,1998:Emissions of volatile organic compounds (primarily oxygenated species)from pasture.J.Geophys.Res.,103,10,605–10,619. Knyazikhin,Y.,J.V.Martonchik,R.B.Myneni,D.J.Diner,and S.W.Running,1998:Synergistic algorithm for estimating vegetation canopy leaf area index and fraction of absorbed pho-tosynthetically active radiation from MODIS and MISR data.J.Geophys.Res.,103,32,257–32,276.

Legates,D.R.,and C.J.Willmott,1990:Mean seasonal and spatial variability in global surface air temperature.Theor.Appl.Climatol.,41,11–21.

Los,S.O.,C.O.Justice,and C.J.Tucker,1994:A global1x1NDVI data set for climate studies derived from the GIMMS continental NDVI data.Intl.J.Remote Sens.,15,3493–3518. MacDonald,R.C.,and R.Fall,1993:Acetone emissions from conifer buds.Phytochemistry,34, 991–994.

Malmstro¨m,C.M.,M.V.Thompson,G.P.Juday,S.O.Los,J.T.Randerson,and C.B.Field, 1997:Interannual variation in global scale net primary production:Testing model estimates.

Global Biogeochem.Cycles,11,367–392.

:A three-dimensional Mu¨ller,J.-F.,and G.Brasseur,1999:Sources of upper tropospheric HO

x

study.J.Geophys.Res.,104,1705–1715.

Olivier,J.G.J.,J.J.M.Berdowski,J.A.H.W.Peters,J.Bakker,A.J.H.Visschedijk,and J.P.

J.Bloos,2001:Applications of EDGAR.Including a description of EDGAR V3.0:Refer-ence database with trend data for1970–1995.RIVM,Bilthoven.RIVM Rep.773301001/ NRP Rep.410200051,159pp.

Potter,C.S.,J.T.Randerson,C.B.Field,P.A.Matson,P.M.Vitousek,H.A.Mooney,and S.

A.Klooster,1993:Terrestrial ecosystem production:A process model based on global sat-

ellite and surface data.Global Biogeochem.Cycles,7,811–841.

Running,S.W.,and R.R.Nemani,1988:Relating seasonal patterns of the AVHRR vegetation index to simulated photosynthesis and transpiration of forests in different climates.Remote Sens.Environ.,24,347–367.

Schade,G.W.,and A.H.Goldstein,2001:Fluxes of oxygenated volatile organic compounds from a ponderosa pine plantation.J.Geophys.Res.,106,3111–3123.

Sellers,P.J.,C.J.Tucker,G.J.Collatz,S.O.Los,C.O.Justice,D.A.Dazlich,and D.A.

Randall,1994:A global1?1NDVI data set for climate studies.Part2:The generation of global?elds of terrestrial biophysical parameters from the NDVI.Intl.J.Remote Sens.,15, 3519–3545.

Singh,H.B.,M.Kanakidou,P.J.Crutzen,and D.J.Jacob,1995:High concentrations and photochemical fate of oxygenated hydrocarbons in the global troposphere.Nature,378,50–

54.

Singh,H.B.,et al.,2000:Distribution and fate of selected oxygenated organic species in the troposphere and lower stratosphere over the Atlantic.J.Geophys.Res.,105,3795–3806. Singh,H.B.,Y.Chen,A Staudt,D.Jacob,D.Blake,B.Heikes,and J.Snow,2001:Evidence from the Paci?c troposphere for large global sources of oxygenated organic compounds.

Nature,410,1078–1081.

Warneke,C.,T.Karl,H.Judmaier,A.Hansel,A.Jordan,W.Lindinger,and P.J.Crutzen,1999: Acetone,methanol,and other partially oxidized volatile organic emissions from dead plant matter by a biological processes:Signi?cance for atmospheric HO

chemistry.Global Bio-

x

geochem.Cycles,13,9–17.

Wennberg,P.O.,et al.,1998:Hydrogen radicals,nitrogen radicals,and the production of ozone in the upper troposphere.Science,279,49–53.

Earth Interactions is published jointly by the American Meteorological Society,the American Geo-

physical Union,and the Association of American Geographers.Permission to use?gures,tables,

and brief excerpts from this journal in scienti?c and education works is hereby granted provided

that the source is acknowledged.Any use of material in this journal that is determined to be‘‘fair

use’’under Section107or that satis?es the conditions speci?ed in Section108of the U.S.Copy-

right Law(17USC,as revised by P.L.94-553)does not require the publishers’permission.For

permission for any other form of copying,contact one of the copublishing societies.