Atmosphere-biosphere exchange of CO 2 and O 3 in the central
Amazon Forest
The Harvard community has made this article openly available.
Please share how this access benefits you. Your story matters.
Citation
Fan, Song-Miao, Steven C. Wofsy, Peter S. Bakwin, Daniel J.
Jacob, and David R. Fitzjarrald. 1990. “Atmosphere-Biosphere
Exchange of CO2 and O3 in the Central Amazon Forest.” Journal
of Geophysical Research 95 (D10): 16851.
doi:10.1029/jd095id10p16851.
Published Version
doi:10.1029/JD095iD10p16851
Accessed
June 15, 2017 9:38:06 AM EDT
Citable Link
http://nrs.harvard.edu/urn-3:HUL.InstRepos:14121810
Terms of Use
This article was downloaded from Harvard University's DASH
repository, and is made available under the terms and conditions
applicable to Other Posted Material, as set forth at
http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-ofuse#LAA
(Article begins on next page)
JOURNAL OF GEOPHYSICAL
RESEARCH, VOL. 95, NO. D10, PAGES 16,851-16,864, SEPTEMBER 20, 1990
Atmosphere-Biosphere
Exchangeof CO2 and03 in the
Central Amazon
Forest
$ONG-MIAOFAN, STEVENC. WOFSY,PETER$. BAKWIN,ANDDANIELJ. JACOB
DivisionofAppliedScience
andDepartment
ofEarthandPlanetaryScience,
Harvard University,Cambridge,Massachusetts
DAVID R. F1TZJA•ALD
Atmospheric
Sciences
Research
Center,StateUniversity
ofNewYorkat Albany
Measurements
of verticalfluxesfor CO2andO3 weremadeat a level 10m abovethecanopyof theAmazon
forestduringthewetseason,
usingeddycorrelation
techniques.
Verticalprofilesof CO2andO3 wererecorded
continuously
fromabovethe canopyto the soil surface,andforestfloor respiration
wasmeasured
usingsoil
enclosures.
Nocturnalrespiration
of CO2 by theforestecosystem
averaged
2.57 kgC/ha/h,with about85 %
fromtheforestfloor. Duringthedaytime,CO2 wastakenup at a meanrateof 4.4 kgC/ha/h.Net ecosystem
uptake
of carbon
dioxide
increased
withsolar
fluxby0.015(kgC/ha/h)/(W
m-2),corresponding
tofixation
of
0.0076molesCO2 permolephotons(about0.017molesCO2permoleof absorbed
photonsat photosynthetically activewavelengths).
The relationship
betweennet ecosystem
exchange
and solarflux wasvirtuallythe
samein the Amazonforestasin forestsin Canada(Desjardinset al., 1982, 1985) andTennessee
(Baldocchiet
al., 1987a,b). The relativelyhighefficiencyfor utilizationof light (about30% of thetheoretical
maximum)and
the strongdependence
of net CO2 uptakeon solarflux suggest
thatlight may significantlyregulatenet ecosystem exchange
andcarbonstoragein thetropicalforest.Changes
in thedistribution
of cloudcover,associated
for examplewith climaticshifts,mightinducegloballysignificant
changes
in carbonstorage.Ratesfor uptake
of03 averaged
2.3x10
ll molecules
cm-2s
4 inthedaytime
(10hours,
700-1700
hours),
dropping
byroughly
a
factor
of 10during
the14hours
fromdusk
todawn.Themean
03 deposition
velocity
at40mwas0.26cms4
inthenightand1.8cms-1 in theday.Diurnal
variation
ofO• deposition
wasregulated
bothbystratification
of
theatmospheric
boundary
layerandby stomatal
response
to lightandwaterdeficit.The totalflux of 03 to the
forestwaslim.ited largelyby supplyfrom the free troposphere
above. Depositionof 03 to the forestcanopy
appears
to bea regionally,andperhaps
globally,importantsinkfor tropospheric
Oa.
1. INTRODUCTION
Depositionratesfor O3 to theAmazonforestwereestimatedfor
tl-,,•
,4..,,
"
....
Tropical
forests
represent
significant
sources
orsinks
forchemi[Gregory
etal.,1988;
Browell
etal.,1988;
Kirchhoffet
ai.,1988].
tallyand
climatically
important
trace
gases,
including
03,CH4,Theanalysis
indicated
a downward
fluxin excess
of 1012
CO2,
andreactive
hydrocarbons.
Rapid
rates
foratmospheremolecules
cm-2s
-1,buttheestimates
could
notaccount
forphoto-
biosphereexchangeare associated
with high temperatures
and chemicalproduction[JacobandWofsy,1988]. Depositionratesat
humidities,
high
rates
ofbiological
activity,
and
intense
sunlight.
night
were
obtained
byKaplan
etal.[1988]
using
measured
vertiRecent
studies
have
shown
that
deposition
tovegetation
isasigni½al
gradients
ofO3within
and
above
thecanopy
and
vertical
exticant
sink
forO3inthe
planetary
boundary
layer
over
thetropical
change
coefficients
derived
from
fluxes
and
gradients
ofNO.The
forest
[Gregory
etal.,1988;
Kaplan
etal.,1988],
and
may
also
be lackofdirect
information
ontrace
gasexchange
over
tropical
animportant
contributor
totheglobal
budget
oftropospheric
03 forests
makes
itdifficult
todetermine
thefactors
that
regulate
gas
[Galbally
and
Roy,
1980;Liu,
1988].
Exchange
ofCO2
intropical
fluxes
andtoassess
theroleoftropical
forests
intheglobal
forests
maylikewise
besignificant
intheglobal
carbon
cycleatmosphere-biosphere
system.
[Fung,
1986;
Matthews,
1983;
Mooney
etal.,1987].
Wepresent
inthispaper
thefirsteddy
correlation
measureDirectmeasurements
of gasexchange
betweentropicalforestsmentsof 03 depositionand CO2 exchangeover a tropicalforest.
andtheatmosphere
aresparse,
reflecting
thedifficulty
ofmountTheeddy
correlation
method
hasthevirtue
ofbeing
direct
and
ingcomplex
experiments
inremote
locations.
Uptake
rams
for non-intrusive,
andresults
maybeobtained
continuously
over
exCO2
byphotosynthesis
have
been
estimated
indirectly
using
varia-tended
periods.
Theexperiment
wasconducted
atDucke
forest
tions
ofCO2observed
within
and
above
thetropical
forest
canopy
reserve
near
Manaus,
Brazil
aspartoftheNASA/INPE
ABLE2b
[e.g.,
Odum
andJordan,
1970;
Lemon
etal.,1970;
Wofsy
etal., mission
during
thewetseason
of 1987.Fluxes
andbudgets
of
1988].
Emissions
ofCO2from
soils,
which
contribute
most
ofthe healwater
vapor,
and
momentum
have
been
studied
previously
at
nocturnal
CO2flux,have
been
measured
using
enclosure
tech-thissitebyShuttleworth
etal.,[1984a]
andFitzjarrald
etal.
niques
IGoreau
andDeMello,
1985;
Keller
etal.,1986;
Schles[1988].
inger,
1977].
Fluxes
ofCO2andOahave
been
measured
byeddy
correlation
over a varietyof othervegetationtypesusingeithertoweror aircraft platforms, e.g. Verma and Rosenberg,[1976]; Ohtaki,
Copyright
1990
bythe.american
Geophysical
Union.
Paper
number
90JD00595.
0148-O227/90/90JD-00595505.00
[1980];
Desjardins
etal.,[1982];
Alvo
etal.,[1984];
and
Baldoc-
chi et al. [1987a,b]for CO2 overpaddyfield andforest; Weselyet
al. [1978, 1981]; Lenschowet al. [1981, 1982]; and Droppo,
[1985] for 03 overforest,water,andothersurfaces.Hicksandhis
16,851
16,852 •
[:AN t:] •1.: (;ARB()N [)I()XII)E &N[) (_)Z()Nt I.X½'IIANGE
ABLE-2B
coworkers
[e.g.,HicksandMcMillen,1988;McMillen,1988]and calibrations
werecardedout at thispressure.
The modifiedCO2
Denmeadand Bradley[1985, 1987] havecontributed
importantinstrument
exhibitedan artificialmodulationat about2 Hz, but
recentadvancesin the applicationof eddy correlationmeasure- thisnarrowband,highfrequency
component
did notcorrelatewith
mentsto forests.Thesestudiesprovidethe background
required verticalvelocityor contribute
to CO2fluxes(cf.nextsection).
to designa measurement
strategyfor tropicalrainforests,includ- Air passedthrougha thermoelectrically
cooledglassvolume
ingprecautions
neededto makemeasurements
in nonidealterrain filledwithglassbeadsbeforeentering
theCO2analyzer,
establishusingsensors
withimperfect
response
functions
andnoise.
ingconstant
temperature
anddewpoint(5øC). Liquidwaterwas
We alsomadecontinuous
measurements
of mixingratiosfor drainedthrough
an openingat thebottomof thevolumeby drawCO2andO3 at eightaltitudes
aboveandwithinthecanopy.The ing off an air flow of 0.3 L/min. 'Thisarrangement
assured
that
profilesprovidemeasurements
of storagein the canopylayer. fluxesdo not haveto be corrected
for density
variations
dueto
Canopystorage
contributes
a significant
partto thenetecosystemfluctuations
of temperature
or watervaporin the samplecell
exchange
atcertain
timesof day,particularly
forCO2.
[Webb
et el., 1980].
Theresults
clef'me
thediurnal
variation
oframsforgasexchange TheO3 analyzer
(MonitorLabs,Model8410)detects
chemibetween
theforestandtheatmosphere,
andprovideinformationluminescence
of thereaction
of O3withC2H,•. It wasmodified
onthefactors
thatregulate
gasfluxes.We examine
theelementsforrapidresponse
byGregory
etal. [1983,1988],attaining
a 90%
thatcontrol
atmosphere-biosphere
exchange
usinga canopy
resis- response
timeof 0.8 seconds.
Calibration
of theinstrument
was
rancemodelsimilarto that usedby Hicksand coworkers
to checked
in thefieldby comparing
withO3 concentrations
measdescribe
deposition
of 03, SO2,andNO2to vegetated
surfacesuredby theDasibiModel1003AHO3 analyzer
at 41 m height
suchasmaizeor oak-hickory
forest[Baldocchi
et el., 1987;Hicks [Bakwinet el., thisissue(a), (b)]. Instrument
response
wasexetel., 1987;Meyers
andBaldocchi,
1988].Themodelanalysis
in- tremely
stable
duringthecourse
of theexperiment.
dicates
thatatmospheric
andbiospheric
processes
arebothimpor- Rawdataforw (theverticalwindspeed),CO2,andO3 wereactantin limitingdeposition
of reactivespecies
to thetropicalforest. quired10 timesper secondusinga PDP 11/73computersystem
andwererecorded
onfloppydiskettes.Spectra,
cospectra,
andco2. EXPERIMENT
variancesof w, CO2, and03 werecalculatedlater. Data weretak-
en in parallelat 1 secondintervalsby a CampbellScientificDaThe experiment
wasconducted
at theReservaFlorestalDucke, mlogger.The datalogger
storedreadingsfor 20-minperiods,then
25 km northeastof Manaus,Amazonas,Brazil, betweenApril 22 removedthe mean for eachsensorand computedvariancesand
andMay 8, 1987. Vegetationand climatologyfor the sitehave covariances,accountingfor time delaysbetweeninstruments
beendescribed
in detail
byShuttleworth
et al. [1984a,b]
and rounded
tothenearest
1 second
[Fitzjarrald
etal.,thisissue].
Roberts
etal. [1990].
Themain
treecanopy
extends
about
30m Timedelays
between
thechemical
andthevertical
windsignals
above
ground
withsome
emergent
trees
to 35 m. Foliage
is were
estimated
inthefieldbyblowing
breath
intotherobing
inlet
present
atalllevels,
though
vegetation
density
ishighest
in the (forCO2)
andbyturning
onandoffanozone
generator
(Hgvapor
upper
portion
ofthecanopy
andnear
theground.
Diurnal
varia-lamp)placed
neartheinlet(for03). Theestimated
delays,
tions
of temperature
andspecific
humidity
deficit
werestrong
at 8.1-1-0.1
sforCO2and10.2-1-0.1
sfor03,wereused
influxcalcutopof thecanopy,
ranging
from22 to30øCand-0 to 8 gtKg lations
throughout
theexperiment.
These
values
wereconf'mned
(night-m-day)
respectively.
Diurnal
variations
oftemperature
and bycomputing
lagged
covariances
fromtherawdataasdiscussed
humiditywerevery smallnearthegroundandthesoiltemperaturebelow.
wasalmost
constant
at25øC.Meancloudiness
isabout
0.6inthe Concentration
profiles
for CO2 andO3 wereobtained
by
dryseason
and0.7inthewetseason
[Ratisbona,
1976].
sequential
sampling
ateightaltitudes
(0.02,3, 6, 12,19,27,36,41
Analuminum
scaffolding
towerof 45 m height
served
asthe m)through
fixed1/4-inch
ODTeflon
tubes.
Theinlets
ofthetop
mainplatformfor theexperiment.The flux measurements
of CO2 five levelswere attachedto rodsextendedaboutI m from the
andO3 wereacquired
at 39 m on thetower,about9m abovethe tower,whileinletsof thelowerthreelevelswereattached
to a tree
canopytop. A fastresponse
single-axis
sonicanemometer
(Camp- about15 m fromthetowerin orderto avoidinfluence
of human
bellScientific
Inc.) wasmounted
attheendof a beamextendingactivityaround
thebaseof thetower.A manifold
of solenoid
about
3 mfromthetower
[Fitzjarrald
etal.,thisissue].
Operation
valves
wasusedto switch
between
inlettubes
forairsampling,
of thisinstrument
and associated
measurements
of heatand fromlowest
tohighest,
witha dwelltimeof4 minateachlevel.
momentum
fluxatseveral
altitudes
arediscussed
byFitzjarrald
et Concentration
profiles
forO3weremeasured
usinga Dasibi
al.[1988,andthisissue].
1003-AHozoneanalyzer
described
by Balewin
et al. [thisissue
A Teflontubeof 3/8"outside
diameter
(8 mmi.d.) wasat- (a), (b)]. Themeasurements
of CO2profiles
weremadewitha
tached
to thebeamabout30 cm fromtheanemometer,
openingBeckman
865 nondispersed
infraredanalyzer.Because
of the
downwards
andprotected
by a smallplasticfunnelto avoidas- variablelengthsof the sampling
tubes,cell pressures
in the
pirating
rain. Ambient
airforanalysis
of CO2and03 wasdrawn analyzer
variedfrom580to600torrsaccording
tothealtitude
bethrough
theTeflontubeata totalflowrateof25standard
litersper ingsampled.
Calibrations
werecarriedoutat several
pressures
minute,smallenough
to avoidinterference
withthemeasurementapproximately
every2 hours.A cooler,identicalto thatusedin
of verticalwindvelocity.Mostof theair flowwasbypassed
just thefast-response
instrument,
conditioned
sample
airandstandards
upstream
of the analyticalinstruments,
with 2.5 L/min drawn toconstant
temperature
anddewpoint.Carbon
dioxide
concentrathxough
theCO2analyzer
and0.84L/minthrough
theO3analyzer tionswerecalculated
by interpolating
betweenconsecutive
gain
(MonitorLabs,model8410).
factorsandzeros,andwerecorrected
for pressure
effects.ProillA BINOS nondispersed
IR analyser
wasusedfor theCO2flux ing procedures
were controlled,and datarecorded,by an HPmeasurements.
The instrument
electronics
weremodifiedto give 3421Adataacquisition/control
system
andanHP-85computer.
rapidresponse
(90% response
tijne4).3 s), so thattheexperimen-
Soil emissionsof CO2 were measuredat sevenlocationsnear
halresponse
timewaslimitedby timerequiredto flushthe25 mL the baseof the micrometeorological
towerusingTeflon coated
sample
cell(0.6s)andcoldtrap(-0.3 s,seebelow).A pressure
of aluminum
enclosures.
The enclosures
weremadein two parts:
about650 tortswasmaintained
in sampleandreferew.
cecells,and roundcollars16.4 cm deepand24.8 cm diameterplaceda few
ABLE-2B
FAN ET AL.: (dARB()N DI()XIDI::. AND OZ()NE
centimeters
into soil, and tops 19 cm high. Each top had two
EX('ttANGE
16,853
!.0
holes, 1.5 cm in diameter,one connectedto 1/4-inch Teflon tube
leadingtogasanalyzers
andtheotheroneopento theair. Theair
inside the enclosurewas mixed gently by an aluminumpaddle
drivenby a batterypoweredmotorat 120 rpm. The collarswere
insertedin the soil at least2 daysprior to the measurements
and
most were left undisturbedthroughoutthe experiment. Samples
were takencontinuously
at a flow rate of 300 mL/min usingthe
BeckmanCO2 analyzerimmediatelyafter a top was put on the
collar. Carbon dioxide emissions were calculated from the rate of
increaseof CO2 concentrationinsidethe chamber.
Total incidentsolarradiationwasroutinelymonitoredby a py-
roheliometer
at a site2 km awayfrom themicromettower(O. Cabrai, unpublisheddata, 1989). We retrievedsolarflux data from
chartrecordsfor 50 daysin April andMay 1987;thedataweredigitized, rectified, and converted to incident flux using the
manufacturer's
•' -2.o
calibration.
I
.4o
3. VERIFICATIONOF FLUX DATA
I
.2o
I
I
I
o
2o
40
6O
SECONDS
The verticalflux of a scalar,suchasCO2 or 03, canbe viewed
as an imbalm•cebetweenthe quantityof scalartransported
across
a horizontalplaneby air parcelsmovingup anddown.The time
averaged
verticalHuxF maybe represented
by
F=(1/r)lw'(Oc'(t+at)at
o
I
(•)
whereT is the averaginginterval,andAt is thedelaytimebetween
themeasured
signalsfor verticalvelocity,w, andconcentration,
c.
The primesin (1) denotedeviations
of w andc from theirrespective nmningmeanscomputed
overintervalsof lengthT. The sensorsfor w and c must be sensitiveenoughand respondrapidly
enoughtoresolvesmalldeviations
at thefrequencies
important
for
turbulenttransport.Averagingintervals(T) mustbe sufficiently
shortto resolveatmospheric
variations
in flux butsufficientlylong
to obtainstatisticalsignificance.A periodof 20 min to ! hour is
oftenchosento obtainan effectiveaverageovera largenumberof
eddies,while allowingobservations
of thechangein theflux with
time[LumleyandPanofsky,1964].
The observedsignalsare affectedby instrumentalartifacts
_111
t_
-20
-60
-40
-20
0
20
40
60
SECONOS
(baseline
driftandnoise)
thatmayproduce
errors
inthecomputed
Fig.1.Lagged
covariances
(a)between
wand
CO2
and
(b)between
wand
flux.Exposure
andorientation
oftheanemometer
mayalso
affectO3calculated
fortheperiod
14-00-1620
onMay2.(Negative
lags
when
thequality
offluxdeterminations.
Shuttleworth
etal.[1984a]
and measurement
ofCO2
and
O3were
delayed.)
Fitzjarrald et al. [1988, and this issue]have demonstrated
previouslythevalidityof eddyflux measurements
for momentum,
heat
andwatervaporat thissite. We examineherein detailtheaccura-
cyof thetracerfluxesdetermined
by theDatalogger,
whichcon- Crosscorrelation
functions
sLrnilar
to t.hatshownin Figure1
stitute
themaincomponents
of ourdataset(onlya fewhours
of werecalculated
forotherperiods.
Thedelaytimes
obtained
from
rawO3datawereacquired
by thePDPsystem
dueto logistical
these
correlation
functions
varyordya fewtenth
seconds
fromone
problems.)
Threesources
of errorwillbeexamined:
(1)errors
in periodto another.
Asindicated
in Figure1, anerrorin thetime
delaytimes
used
bytheDatalogger,
(2)instrumental
zerodriftand delayof0.1s would
introduce
anerrorin fluxof about
1 percent
selection
ofdataaveraging
interval,
and(3)statistical
imprecision
forbothCO2and03,small
compared
toother
sources
oferror.
associated
withfluctuation
of transport
ratesandwithaliasing
of Instrumental
zerodriftrepresents
lowfrequency
noise.It can
high-frequency
noise
fromthesensors.
bea serious
source
of errorif zerodriftoccurs
forbothw andc
Timedelays
of 8 s forCO2and10s forO3 wereused
bythe signals
andaveraging
does
noteffectively
remove
itscontribution
Datalogger
throughout
the experiment.
Laggedcovariances
to thefluxestimate.
Thefluxeswerecalculated
fromtheDabetween
w andCO2andbetween
w andO3werecalculated
usingta!ogger
measurements
bysubtracting
20-minmeans
fromw and
10Hz PDPdatato investigate
thesensitivity
of computed
fluxesc, thenperforming
theintegral
in (1). In ordertoassess
themagto deviations
fromtheassumed
values.Themagnitude
of theco- nitudeof errorsintroduced
by theDatalogger
algorithm,
wecomvariance
of w' andc' should
be maximum
at thecorrect
delay putedfluxesof CO2and03, using10Hz rawdata,fora series
of
time.Examples
of thecovariance
asa function
of delaytimeare running
meanintervals
from1.7to 20 min. Results
(Figures
2a
shown
in Figuresla andlb, in whichoptimalcorrelations
were and2b) indicatethatthe computed
flux is independent
of the
foundat 8.5 s between
w andCO2and10.2s between
w andO3, averaging
intervalfor intervals
between
6 and20 min. The comin goodagreement
withestimates
madein thefield.
putedflux declines
for intervalslessthan5 minutes.When20-
16,854
FAN ET AL.' C'ARB()N DIOXIDE AND O7_.()NEEX('i-tANGE
ABLE-2B
C02 FLUX
lO
5
-4
I'
I
I
I
,I
0
5
10
15
20
AveragingInterval(min)
L,
1
03 FLUX
0.005
I
I
I
I
I
0.01
0.05
0.1
0.5
1
f(hz)
5o
-lO
10
- 2o
5
-3o
1
0.5
-4o
ro
I5
I-
lO
i
15
i
20
AveragingInterval(min)
0.i
0.05
Fig. 2. Fluxesof (a) CO2 and(b) 03 calculatedwith runningmeansof differentlengthssubtract•fromsignals.Dataarefrom 1400-1640,May 2.
0.01
•
•
I
I
I
I
0.005
0.01
0.05
0.1
0.5
1
min grandmeanswereremovedfrom signalsaveragedfor the 16min period(1400-1640)on May 2, 1987,fluxeswereat most15%
differentfrom thosecomputedby subtractingthe 6 min running
mean(seeFigure2). Theseresultsweretypicalof othertimeintervals,asindicatedby Figure4 (discussed
below). The smallerrors associatedwith zero drift reflect stableinstrumentperformancein enviromental conditionsthat variedlitfie duringthe
period of experiment.We concludethat subtractionof running
meanoverintervalsbetween6 and20 min, or a grandmeanover a
20-min
period,
provide
fluxestimates
consistent
tobetter
than
15% for most intervals.
Statistical errors in estimates of the mean flux are associated
0.5
with temporalvariationof the verticalflux and with instrumental
noiseproducedby fastresponse
chemicalsensors.Statisticaliraprecisionis usuallythe largestsourceof errorfor a flux estimate
in a single20-minuteinterval,typicallyat least10%underideal
conditions[Weselyand Hart, 1985].The errorsin a givenperiod
dependon thevariancesof w andc, andon the signal-to-noise
ratio of eachmeasurement[Lenschowand Kristensen,1985].
Figures3a-3c showpowerspectrafor w, CO2, andO3 sensors,
computedusinga Fouriertransformof 10-Hz datafor a 140-min
0.1
0.005
I
!
I
I
I
0.01
0.05
0.1
0.5
1
5
f(hz)
interval
in theafternoon.
Thespectra
forCO2andO3havesigni- Fig.3. Examples
of power
spectra
andcospectra
atarbitrary
units.(a)
ticantpowerat highfrequencies
thatis notobserved
in thespec- Spectrum
ofw,(b)spectrum
ofCO2,(c)spectrum
ofO3,(d)eospeetmm
of
tramfor w. HoweverFigures3d and3e showthatthecospectxa
of w withCO2,and(e)cospectmrn
of w with03. Dataarefrom1400-1620,
w with CO2 and O3 showno significantcontributions
from fre, May2.
ABLE-2B
FAN El' At..: CARB()N [)!½)Xil_)l•AND OZ()NE EX(•HANGE
16,855
data (+ connectedby lines) for the intervalswhen both systems
were operating.Fluxesderivedfrom 10 Hz datavary smoothly
over periodsof severalhours. The fluxes obtainedfrom the Damloggerare scatteredaboutthe smoothlines,but the magnitude
0.0
and diurnal variation
-0.5
of fluxes derived from the 10 Hz data are
well reproduced.Errorsfor hourly averagedfluxes are in most
caseslessthan-20% of peakvalues.Comparison
of hourlyfluxes
from the Dataloggerto thosecomputedfrom 10 Hz data (Figure
4c) indicates
a regression
(r2=0.66)witha slopeverycloseto 1
and with zero intercept. The residual standarderror is 1.9
kgC/ha/h,with no systematictrendin theresiduals.
The influenceof aliasingby the Dataloggeralgorithmwas
furtherexaminedby computing
CO2 and03 fluxesfromPDP data
sampledonce per secondto simulatedata collectionby the Datalogger.Mean fluxesandstandarddeviationsareshownin Table
1 alongwith the corresponding
fluxesfrom the Datalogger.The
errorsintroducedby the low samplingrate areevident.The coef-
-2.0
-2.5
-3.0
0.005
0.01
0.05
0.1
I
I
0.5
1
5
f(hz)
ficient of variation for O3 flux is smaller than that for CO2, reflectingthehighersignal-to-noise
ratio of the ozonemeasurement.
Nevertheless,fluxes averagedover eight intervalsfrom the Dataloggerand the PDP agreecloselyfor both 03 and CO2, consistentwith our analysisof errorsfor the techniques.
Data were obtainedfor varioustime intervalson 12 days, and
for each hour we have from the Datalogger19-29 (mostly27),
20-rain intervalsfor CO2, and 24 20-rain intervalsfor 03. Accordingto the analysisof errorsoutlinedabove,randomerrorsin
hourly mean values for the whole experiment,due to statistical
variance,removalof means,etc., shouldbe lessthan5% of daily
maximafor mosthoursof the day. As we shallseebelow, biasin
the data due to variations of natural conditions(cloudiness,wind
speed)andby floodingof the anemometer
duringrainstomsmay
introducelargererrorsin definingmeanvaluesfor the experimental period.
-5
4. CA."_nONDIOXIDE FL?X .s.N• N•
I
0.O05
0.01
o .o5
Fig. 3.
I
o.1
I
0.5
(continued)
F.,cosYs,•,¾, PRODUCrI,,',,,rrl'Y
A characteristic
diurnalpatternwas observedfor CO2 flux (12
daysof data)anda verysimilarpatternwasobservedfor 03 (partial datafor 6 days),as shownin Figure5. Concentrations
above
and within the canopyexhibitedmarkeddiurnalpatterns(Figure
6). The meandaily minimumfor CO2 abovethe canopywas340
ppm at midday,7-8 ppm lowerthanthemeanvalueobservedat 3
km altitudefrom theElectraaircraft[R. HarrissandS. Wofsy,unpublisheddata, 1989]. The meandaily maximumat canopytop,
more than 370 ppm, was observedas CO2 accumulatedin the
quencies
above1 Hz; evidentlyspurious
high-frequency
signals forestboundary
layerbeforesunrise.Variationof totalCO2condonotcorrelate
wi•hw sufficiently
to introduce
significant
errors tentbetween0 and39 m heightis relatedto thefluxesat thetop
intothefluxcalculation,
atleastwhenthesignalsareoversampledand bottomboundaries,
and to production
and losswithinuhe
by acquiringreadingsat 10 Hz, high-frequency
noisedoesnot coltombby
contribute to the mean flux since it does not correlate with w.
39m
'•dI c(z)dz
Most
ofthe
power
inthe
cospectra
inFigure
3corresponds
to
+F39m
=F•oa+F•,f(2)
periods
between
10 and100s,longerthanwouldbetypicalover
low-stature
vegetation[Finnigan,1985].The cospectra
for w with
CO2andOa,andforwwithtemperature
andwater
vapor
[Fitzjar-where
F•oa
istherateofemission
bysoils,
F•oa
istheintegrated
raidetal.,thisissue]
'indicate
thatcontributions
tothefluxfromrateofemission
byleaves,
andF39m
istheupward
fluxat39m.
frequencies
above
1Hzwere
negligible.
It istherefore
notnecesTheinfluence
ofhorizontal
advection
isassumed
tobenegligible
sarytocorrect
computed
fluxes
forlimited
bandpass
ofthetraceonaverage.
Theleft-hand
side
in(2),thenetecosystem
exchange
gassensor
systems
[cf.Hicks
andMcMillen,
1988].
(NEE),canbecomputed
using
measurements
ofvertical
concenSpurious
high-frequency
signals
affect
theDatalogger
algorithm
tration
profiles
andfluxatthetopofthetower.
more
than
thePDP
fluxdetermination,
because
inthis
case
the Figure
5compares
thediurnal
variations
ofstorage
(d/dr
1309•
signalsare not oversampled.A set of data acquiredonceper c(z)dz)withthemeasured
meanfluxes(F39m).Diurnalchanges
in
secondcan alias high frequencyvariationsinto the flux signal. storagetermswere significantfor the budgetof CO2, but not for
Figures4a and4b showhourlyaveragedCO2 fluxesderivedby 03. Storagetermswere particularlyimportantduringthe early
the Datalogger(squares)andfluxesderivedby analysisof 10 Hz morninghours,whenCO2 respiredat nightis usedfor photosyn-
16,856
FAN ET AL' CARB()N DI()XIDE AND OZ()NE EX(;}iANGE
ABLE-2B
10
aaa t
a
•
aaaa
o• ßa
•
-- aaaaa
D
D
D
D
D
./
12
-10
1"
!
I
!
..!
0
20
40
60
80
HOURS
10
APR 27-29
_
o1:I a
a
rl
'5
ß'
D
a
.
-10
-15
1"
,
0
!
.!
20
40
HOURS
'
I
•
60
80
MAY 6-8
Fig. 4. Comparison
betweenhourlyCO2fluxes(kgC/ha/h)fromtheDatalogger
andPDP systems.
(a) Time seriesfromApril 27
to April 29 (squares,
Damlogger,crosses
connected
by linesfor PDP). (b) Time seriesfrom May 6 to May 8 (samelabelsasin
symbolsFigure4alfR). (c) Datalogger
flux versusPDP flux (squares),
regression:
¾= 0.97 (i0.09, standard
deviation)
X = 0.0
(i0.25,standard
deviation),
r2=0.66.
thesiswhilesimultaneously
beingexhausted
fromtheforestto the
Themorning/evening
(am/pm)ratiofor CO2uptake(1.31)was
developing
atmospheric
mixedlayer.
virtuallythe sameas the am/pmratio for incidentsolarflux (see
The maximumuptakeof CO2 by the forestsystemoccurred Figure8), reflectinggreatercloudiness
for afternoon
sampling
in-
around
noontime,
onaverage
about9.3kgC/ha/h.
Meandaytimetervals:themeansolarfluxwas460W m-2 in themorning
and
(6-18hours)
NEEforCO2wasabout-4.4
kgC/ha• withgreater360 W m-2 in theafternoon
(am/pm=1.28).
Neithertheasymuptakeobserved
beforenoon(-5.0kgC/ha/h)thanin theafternoonmerrynorthemagnitude
of thesolarirradiance
weretypicalof the
(-3.8 kgC/ha/h).The flux of CO2 from the forestfloor averaged 50 daysin April andMay 1987,for whichwe obtainedsolarflux
2,22 kgC/ha/h,with little variationaccording
to timeof day(Fig- data: averagevaluesfor the solarflux in themorningandafter-
ure7). Thenighttime
meanvaluefor NEE was2.57kgC/haPn,
noon were 320 W m-2 and 325 W rn-2, respectively
alsowithlittlevariationthrough'thenight(Figure5a).
(am/pm=0.99).Intervalswith flux data are evidentlybiasedto-
ABLE-2B
FAN ET At..: CARB½)N
Dit)X!I)E ANt) ()Z()NE EX½'tt,X,
NGE
16,857
002 FLUX- kgC/ha/nr
lO
L
O
G
G
E
R
-5
•
Døa . D
-lO
-5
0
5
10
PDP
y = 0.97 (-,-0.09)x + 0.0 (4-0.,25);R = 0.8
Fig. 4.
TABLE 1. Fluxesof CO2 and03 Recordedby the
C/ha/h,and ca--411(:t:l)(wmZ),(r2=0.90). The averagehalf-
Datalogger
andCalculated
fromthePDPdata
saturationvalue for the forestsystem(ca) is l-dgherthan the day-
COz, kg C/ha/h
Data-
Time
(continued)
O.½I()tl molecules/cm2/s
timemeanof 389W/m2, implying
thatthesystem
is lightlimited:
net •,,•.y
A•;1,,
Data-
!o!•6er Mean* o'f
lo•6er Mean*
of
1420
1440
1500
1520
1540
1600
1620
1640
-6.29
-5.54
-6.02
-4.73
-6.89
-3.15
2.69
-5.68
-3.75
-3.29
-5.39
-7.24
-5.55
-2.41
-4.74
-4.64
0.47
0.39
0.33
0.89
1.33
0.34
0.31
0.34
-2.53
-3.18
-3.73
-4.82
-3.13
-4.38
-3.68
-1.53
-1.64
-2.10
-3.06
-5.72
-4.05
-2.66
-2.93
-3.73
0.08
0.11
0.09
0.14
0.17
0.07
0.11
0.08
Mean
Standard
-5.12
1.50
-4.63
1.50
0.55
-3.37
1.04
-3.23
1.28
0.10
* Meanfluxcalculated
for theintervalby sampling
10Hzdataat 1 Hz and
computing10 valuesof the flux.
* Standarddeviation.
tlAaI.
tUIt •..1•1 v a&"y
Wihh
carbong.'•.,:^_
Analysisof airborneflux datafor the Laroseforestin Canada
[D½sjardins
et al., 1985] gave a relationshipremarkablycloseto
that foundhere for the Amazonforest. They fit a line to uptake
data as a functionof solarirradiance,obtaininga slopeof -0.014
(kgC/ha/h)/(W
m-z) andan'intercept
of about1.2kg C/ha/h.If
we treatthe Amazondam the sameway (see the dottedcurvein
Figure9), weobtaina slopeof -0.015(kgC/ha/h)/(W
m-z) andintercept
of 2.2 kgC/ha• (rZ=0.86).Thecoefficient
forcarbon
uptake observedin a Tennesseeforestby Baldocchiet aL [1987(a)]
wasabout-0.028
kgC/ha/•(W m4 of PAR),andsincePAl! accountsfor abouthalf of the solarflux [Canwbell,1977; Od•m
aL, 1970], thevariationof NEE with solarflux alsocorresponds
to
about-0.014
kgC/ha/h(W
Themaximumquantum
yieldforphotosynthesis
in leavesof
species
is about0.06 moleCO•.fixedper molephotonsat 25øC
wards
sunny
periods,
a consequence
of repeated
failures
of the andintercellular
concentrations
of CO2andO2of 230ppmand
sonic
anemometer
during
rainstorms.
21%,respectively
[Farquhar
etal.,1980].Themean
coefficient
Figure
9 shows
thehourly
mean
NEEforCO2observed
at forCO2uptake
bytheAmazon
forest,
0.015
(kgC/ha/h)/(W
mZ),
Reserva
Ducke,
plotted
against
the'incident
solar
radiation
atthe isequivalent
to a quantum
yieldof 0.0076moleCOz/mole
phoEmbrapa
site2 km away.Carbon
dioxide
is takenupwithre- tons,
using
theconversion
factors
givenbyCampbell
[1977].The
duced
quantum
efficiency
atthehigherirradiance
levels.A least albedo
overthe,aanazon
forestwasestimated
byShuttleworth
et
squares
fit ofthethedatatothefunctional
form
al. [1984a]tobeabout12%,andif PARis assumed
toaccount
for
NEE = ct +
cz'S
50%ofsolar
irradiance,
thequantum
yield
forabsorbed
PARat
ca+ S
Thisis 30%of thetheoretical
maximum.
Forirradiance
levels
themeanirradiance
levelis about0.017moleCOz/molephotons.
between
zeroand200W/m2, thecarbonfixationrateincreases
by
whereNEEis thenetecosystem
exchange
forCO2andS isthein- about0.051molesof COzfor eachadditional
moleof absorbed
cidentsolarflux, gavect= 4.1 (:t:l)kgCPna/•c2=-18.4(:!:3.8)kg
photons
in the PAR spectrum;the marginalquantum
yieldfor
16,858
FAN E'I AL.: CARBON DI()XIDE
AND OZ()NE
EX('IIANGE
ABLE-2B
FLUX(0)STO'RAGE
(+)
2
x
:3
-4
0
o
-6
'-8
-10
o
5
lO
15
20
ß
LOCAL TIME
FLUX(u)STORAGE(+)
.
•
0
.
• .olo
• ..21:)
o,o_.
'•' -30
I•-40
0
'
'5'
•
10'
'
ß•
15
•
20
...
ß
LOOAL TIME
Fig. 5. Meanhourlyfluxesmeasured
at 39mheight(squares)
compared
withchange
in storage
(crosses)
calculated
frommean
concentrations
measured
ateightaltitudes
from2 cmto 41 m aboveground.
TheNEE (plaincurves)
is thesumof flux andchange
in storageß
(a)CO2(kgC/ha/h)
and(b)03 (1010molecules
cm-2 s-l).
CO2 uptakeat low light levelsis thereforenearlyequalto the ecosystem
uptakeof carbon,atleastovershorttimeperiods.
Light
maximum
expected
forC3plants.
sat,marion
effectsarenotdominant
because
theforestis optically
The remarkable
coincidence
of resultsfor boreal,temperate,deep,andmostof theleavesfunction
at lightlevelsbelowsaturaandtropicalforestsstronglysuggests
that,in a forestwith well- fionandat temperanares
withinlimitingvalues.Thehighefficiendeveloped
canopyandadequate
water,availablePAR controlsnet cy for utilizationof light by well-wateredforestecosystems
is
ABLE-2B
FAN E•! a,i•.: CARB()N
Dif)XIDk
AND OZ½)NE EX('tlAN(jE
CO2 CONCENTRATIONS
(ppm)AVERAGEDOVER12 DAYS
16,859
SOIL CO2 FLUX
õ0
.
A 40
,
• -
3O
T
u
E 20
c•
o
'
(M)
1.0
r
1
0
25
0
Fig.
Soil e
7.
•n•
tduring
the
5 wetseason
of
5O
40
3O
t
20
{m)
,"
I
I
i
I
I
I
4
6
8
10
12
14
16
18
20
LOCAL TIME
0
i
i
i
i0
5
10
15
20
25
Local Time
Fig. 8. Incidentsolarradiationaveragedover 50 days (plain curve)and
overperiodswhenthe NEE of CO2 was measured(curvewith squares)in
April andMay.
Fig. 6. Diumal variationsof the mean vertical distributionsof (a) CO2
(ppm)and(b) O3 (ppb).
bon,morethan1.2x10
•2 kgC/yroverthe5x106km2 of theAmazon Basin. This flux may be balancedin part by emissionsof
consistent
withthisview. It mightbepossible
to usethecoeffi- volatilehydrocarbons.
Zimmerman
et al. [1988]estimated
that
cientforNEEversus
solarfluxasa diagnostic
tool,i.e.,if a sys- emissions
of isoprene
andterpenes
shouldaccount
for 0.010and
ternweresignificantly
lessefficient,
another
limitingfactor(e.g., 0.002kgC/ha/h
duringthedryseason,
respectively,
based
onexwater)mightbesuspected.
tensive
measurements
in theplanetary
boundary
layer,Jacoband
Net dailyuptakeof CO2wasobserved
to be 0.93kgCflna/h
for Wofsy [1988] obtainedsimilar resultsfor isoprene(0.016
the periodof experiment,
representing
a significantimbalancekgC/ha/h)usinga photochemical
modelandaircraftobservations
betweencarbonfixationandmineralization.
Thisimbalance
is a of isopreneconcentrations.
Emissions
of reactiveterpenoid
hyfunctionof the samplingbias towardsunnyintervals.The 12- drocarbons
alonemaythusaccountfor 5-6% of theestimated
net
hour-mean
solarflux for intervalswithflux dataexceeded
the50- carbonuptake. Carbonmonoxide[Kirchhoffet al., this issue],
daymeanby 90 W rn-2, equivalent
tostimulation
of CO2uptakemethane,
andparticulate
anddissolved
organic
carbon
arealsoexby 0.67kgC/ha/h(averaged
over24 hours)according
to theresults ported from the forest, and may accountfor much of the
in Figure9. We wouldexpectthereforethat,overthe50 daysin remainder.
April andMay for whichwe havesolardata,netuptakeof CO2 Theresponse
of NEE to variations
in solarirradiance
suggests
wouldbe only0.25kgC/ha/thcloseto thedetection
limit for this thatimportant
shiftsin globalcarbonstorage
couldbe induced
by
experiment.
changes
in thedistribution
of cloudiness,
duefor exampleto cliThe small residualuptakeof CO2 duringthe wettestmonths, mate fluctuationssuch as the E1 Nino-SouthemOscillation. ff
0.25 kgC/ha/h,corresponds
to a globallysignificant
flux of car- cloudcoverincreased
by 10%,(insolation
decreased
by about35
16,860
FAN EI AL.: CARBONDIOXIDEAND OZONEEX('HANtiE
ABLE-2B
3.0
CO;, NEE'
2.$
2.0
I I
•,,-10
-15
0
200
I
400
....
....
I=:
....
I
600
0.0
800
I
0
5
10
15
LOCAL TIME
Wlm2
Fig.9. TheNEEofCO2versus
insolation.
Ourmeasurements
(squares)
Fig.10.Diumal
variation
ofozone
deposition
velocity
(cms-1)measured
andcurvilinear
regression
(solidline) are shown,alongwith a linearfit in theAmazonforest.
(dottedline) compared
in the text with the treatmentof borealforestdata
by Desjardinset al. [1985].
takearethe high densityof vegetation,
thehightemperature
and
humidity(favoringstomatalopening),and rapid turbulenttran-
Wm-2)wemight
expect
net
release
of0.25
kgC/ha•
(24-h
basis)
sport
atcanopy
top.
These
various
factors
are
examined
below.
The downwardflux of O5 measuredat 39 m reflectscontribuor
about
1.2x10
a2kgC/yr.
Although
this
isonly
afactor
of4 fiGns
from
deposition
tovegetation
and
tothe
ground,
and
chemismaller than the annual release from fossil fuel burning, it
cal reactions in the air column between zero and 39 m. Chemical
represents
loss
ofonly
about
0.5%
ofthestanding
biomass
[Medireactions
arenegligible
contributors
tothefluxinthedaytime
na
and
Klinge,
1983]
per
year
and
might
be
sustained
for
a
signifiwhen
the
ozone
flux
is
large
[Jacob
and
Wofsy,
this
issue],
butat
cant period. A decreaseof cloudcovermight stimulatecarbon
night the contributionfrom reactionwith NO emittedby soil is
storage,
although
enhanced
evaporation
might
induce
stress
thatsignificant
[Bakwin
etal.,this
issue
(a),(b).Theflux
F39m
can
be
counteracted
theinfluence
ofhigher
solar
flux.
written
as:
5. FACTORS
CONTROLLING
03 DEPOSITION
rr [O3](z)
[O3](1)
The
daytime
(0700-1700
LT)
fluxes
O3
to
the
forest
canopy
Fag=
=I Rt.(z)
' dL(z)
+I k[NO](z)[O3](z)dz
+ Rs (4)
averaged
2.3x10
xxmolecules
cm-:
s-1,aof
rather
large
value
considering the low O5 concentrations
at canopytop (=6 ppb). Mean
nighttime
fluxes
(1700-0700
LT)werelower
bymore
thanafactorwhere
zr = 30m isthetopof thecanopy,
Rt.(z)istheleafresis-
of 10,2.2x10
xømolecules
cm
-• s-l, while
O5concentrations
at rance
todeposition
persquare
centimeter
ofleafarea,
L(z)isthe
canopy
topwerelowerby onlya factorof 2. Uptakeof O3by leafareaindexbetween
z andzr,k istherateconstant
fortheteac-
vegetation
isevidenfiy
much
more
efficient
during
thedaythan
at fionNO+ 05,andRs istheresistance
todeposition
attheground
night.Thedaytime
O5 fluxesareabout
a factor
of 6 lowerthan based
ontheO5concentration
([O3](1))
at I m altitude.
Theleaf
themid-morning
average
estimated
by Gregory
et al. [1988]for resistance
RL canbe decomposed
intocontributions
fromleafthedryseason.
Thisdifference
canbeexplained
inpartbylarger atmosphere
boundary
(R•), stomatal
(R,),mesophyllic
(R,•),and
O3concentrations
in theplanetary
boundary
layer(PBL)duringcuticular
(Re)resistances,
as follows[Meyers
andBaldocchi,
thedryseason,
about20 ppbin thedaytime
[Gregory
etal., 1988]. 1988]:
The O5 flux F, at altitudez maybe assumed
proportional
to the
concentration
[Oa](z) at that altitude[Galballyand Roy, 1980],
wheretheproponionality
constant
definesthedeposition
velocity
F: = [O3](z)V,•
[
RL=Rb+
•c'c
+ D•
]-'
(5)
(3) where
D andD• arethemolecular
diffusivities
for03 andwater
vapor,respectively.The term representing
reactionwith NO is
We showin Figure10 thediurnalvariationof V,•at 39-maltitude, notstrictlycorrectin thedaytime,sincesomeof theproductNO2
computed
fromthehourlymeaneddycorrelation
fluxesandO5 mayphotolyze,
butthetermis negligible
exceptatnight.
concentrations.
Deposition
velocities
averaged
1.8cms-x in the Weapproximate
theintegrals
in (4) by subdividing
theatmodaytime
(0700-1700
LT) and0.26cms-x atnight(1700-0700
LT). sphere
between
zeroand39 m intodiscrete
layersof uniform
The daytimedeposition
velocitiesare at the highendof values vegetation
densityanduniformO3concentration:
previously
reportedfor vegetation
surfaces
[GalballyandRoy,
"
1980;
Sehmel,
1980].
Inparticular,
Lenschow
etal.
[1982]
meas- F3•m
=•c=•[O:•]•,AL,
R• I+•k[NO]i[Oa]•AZi
+[0:•]
Rs (4')
•x
ureda deposition
velocity
of 1.0cms-x overa GulfCoast
pine
forest,usingtheeddycorrelation
technique.
It appears
thatO5up-
takebytheAmazon
forest
isparticularly
efficient
compared
to where
[Oa]i,[NO]t,andRatrepresent
quantities
averaged
over
otherenvironments.
Possiblefactorsdetermhn.
ing thisrapidup- layeri, AL•is theleaf areaindexof thelayer,andAZ•is thelayer
ABLE-2B
FAN EI AL.: CARBON D!t•Xiœ)E AND OZONE EX(TIANGE
thickness.Inspectionof canopyarchitecture
and of verticalO3
concentration
profilessuggests
a subdivision
of thecanopyinto3 Time
layers: 0-2 m (layer 1), 2-20 m (layer 2), and20-30 m (layer3).
TABLE 2. LeafResistances
(s/m) Usedin theModel
Layer
1consists
ofundergrowth
withhigh
vegetation
density
and 0-6
6-7
low lightlevels,layer2 consists
mainlyof treenunksandhasthe 7-8
lowestvegetationdensity,andlayer3 includesthecrownsof trees, 8-9
which interceptmost of the incominglight. The atmosphere 9-10
above
thecanopy,
between
30and39m,isdescribed
asa fourth10-11
11-12
layer(AL4 = 0), where
O3maybeconsumed
atnightbyreaction12-13
with NO. The concentrations
of NO and O3.usedin (4•) are the 13-14
mean values measuredat the site as a function of time of day 14-15
[Bakwin
etal.,thisissue
(a),(b)].Theground
resistance
Rsisap- 15-16
proximated
astheaerodynamic
resistance
between
0 and1mob-16-17
17-18
rained
fromCO2 andradon
data(Rs=ls cm4 atnight,
and2.5s 18-24
cm4 in thedaytime[Trumbore,
1988].Thesurface
resistance
of
16,861
30-20 m
20-2 m
2-0 m
2.59
1.00
0.75
0.69
0.60
0.69
0.71
0.69
0.78
0.84
0.87
1.07
1.34
2.59
5.27
3.20
2.34
1.95
1.45
1.57
1.76
1.82
2.07
2.13
2.60
2.97
5.10
5.26
11.70
11.36
11.36
7.56
5.85
5.59
5.95
5.59
5.95
5.99
5.84
7.16
11.36
11.70
Computedfrom stomataland leaf boundaryresistancesmeasuredby
the groundis assumednegligible.
Robertset al. [1988] as functionof altitudeand time of day, and distribuThe leaf areaindexof theforestcanopywasnotmeasured
dur, tionofleafareaindex(LAI)discussed
in thetext.
ing the ABLEYB experiment,but previousirradiancemeasurements at the site [Shuttleworthet al., 1984b] indicatethat about
1.4%of thelight at canopytopreaches
thegroundbetween0900
40
and 1000LT (the time of day whencloudiness
is leastfrequent).
Assuming
a uniformangulardistribution
of leaves,corresponding
..
.
,
'
to anextinction
coefficient
of 0.5normalized
toleafareaindex
[Verstraete,
1987],
we
f'md
that
the
light
penetration
observed
by
Shuttleworth
et al. [1984b] corresponds
to a leaf areaindexof 7
(taking
intoaccount
thecorrection
forzenith
angle,
0 = 38ø at • 20.
0930LT). We assme(1,2,4)asthedistribution
of leafareaindex
in thecanopy,wherethenotationrefersto leaf areaindexin (layer
1,layer
2,layer
3)ofthe
canopy
from
bottom
totop;
sensitivity
2.10
calculationswill be conductedwith distributions(1,1,5) and
(1,3,3).
'-
Theleafresistances
Rt.icanbecomputed
from
knowledge
of
theindividual
contributing
resistances
in (5). Leaf-boundary
(Rb)
andstomatal
resistances
(Rs)weremeasured
atthesitebyRoberts
ß
r•
i
1.
,
'
i
i
I
0
LOCAL
TIME
eta!. [1990] Values for R• varied w;,h o•,;,,a•, •om i '• s/cm
neartheground
to 0.3 s/cmat canopy
top;values
for R, varied Fig.11.Uptake
ofozone
byforest
canopy
asmeasured
at39m(squares)
bothwithaltitude
andtimeofday,reflecting
stomatal
response
to andasestimated
bymodel
(lines
frombottom
totop,cumulative
uptake
illumination
andother
parameters.
Theminimum
stomatal
resis-rates
from
ground
through
layers
totopofcanopy).
rancein mid-morning at canopytop was about 1.5 s/cm. We
adoptedaveragevaluesof theresistances
measured
by Robertset
al.[1990]foreach
layerandtimeofday.Mesophyllicresistances
resistance
(Table2), associated
withpartialstomatal
closure
wereassumed
negligible
[Wesely,
1989].A uniform
cuticular
caused
most
likelybywater
stress
[Roberts
etal.,1990].
resistance
R, = 10scm
4 wasadopted,
asdetermined
from
ob- Ouranalysis
shows
thatthedaytime
removal
ofO3byvegetaserved
O3fluxatnight.Thetotalleafresistances
foreach
layer,fioncanbecomputed
using
observed
stomatal
resistances.
GenRt•a• aresignificanfiy
lowerin thedaytime
thanatnight,
as eralizafion
of thisapproach
to other
environments
is however
shown
inTable2,duetoopening
ofthestomata
during
theday.
hampered
by thescarcity
of datafor stomatal
resistances.
A
We compare
in Figure11 theO3 fluxescomputed
from(4•) to numberof modelshavebeenproposed
to predictthe functional
theobserved
values.Reasonable
agreement
is foundthroughout
dependences
of stomatal
resistances
Onenvironmental
variables,
theday.Deposition
totheupper
canopy
(20-30m) accounts
for forvarious
vegetation
types;
a particularly
elaborate
modelhas
about75%of thetotalflux, dueto thehighvegetation
densityin beenproposed
by Baldocchi
et al. [ 1987a,b] whichincludes
the
thatlayerandthelowstomatal
resistances.
Thedistribution
oi?effects
ofphotosynthetically
active
radiation,
temperature,
vapor
leafareaindexin thecanopy
hasonlya smalleffecton thecom- pressure
deficit,•and
leafwaterpotential.Figure12 showsstomaputedflux: the 24-houraverageflux is 11% higherwith the tal resistances.
calculatedusingthe modelof Baldocchiet al.
(1,1,5) distribution,and 11% lower with the (1,3,3) distribution. [1987a, b], adoptingenvironmental
data for the /Xanazon
forest
The agreement
betweenobservedandcomputed
fluxesat nightre- andmodelparameters
for stomatalresponse
derivedby theseau,
flectsmainly the adjustmentof the cuticularresistanceto 10 s thorsfor an oakforest. Watervapordeficitswerecomputedfrom
cm4, a valuemuchlowerthanpreviously
estimated
fordeciduousdataon air temperature
andhumidity.Leafwaterpotential
was
forests[MeyersandBaldocchi,1988;Wesely,1989]. Despitethis assumedto remainhigh throughout
the day. Computedstomatal
relativelylow cuticularresistance,
the daytimeflux is still much resistan• agreewell with observations,
exceptnearthe ground
higherthanthenighttimeflux because
of thelowerleafresistancewherethe modeldoesnot predictsufficientstomatalareaat the
in theuppercanopy(factorof 4) andthe higherO3 concentrationlow ambientlight intensity.The modelsuccessfully
simulatesthe
at 39 m (factorof 2) duringtheday. The decrease
in theflux from increasein resistance
betweenmorningand afternoon,indicating
morningto afternooncanbe explainedby an increase
in stomatal goodrepresentation
of theeffectof waterdeficit
16,862
I-AN E'I At..:CARBf)N[)I{)XII•I::AND()Z()NE !-:.X('ttANGI•:
al., thisissue]suggest
thatthesupplyof O3 to thecanopymaybe
ultimately]knitedby therateof downward
transport
fromthefree
troposphere,
althoughinterpretation
of Lhese
gradients
is complicatedbypotential
photochemical
production
orlossof 03.
Thefactorsregulating
thefluxof O3 to theforestmaybeelucidatedby studying
thesensitivity
of theO3 flux to valuesforcanopy surfaceresistance
or rate of verticaltransport,
usingtheonedimensional
photochemical
modelof Jacoband Wofsy[thisissue]. The model simulatesmeteorologically
undisturbed
conditionsobserved
frequently
evenduringthewetseason,
withdiurnal
growthanddecayof a well-definedmixedlayer. Depositionto
0-2m
10
I
i
I
, i,
8
10
ABLE-2B
i
i
12
14
leavesandgroundis treatedwith the multi-layermodelpresented
above,and verticaltransportis describedby a seriesof aerodynamicresistances
extending
from the groundto the topof the
planetary
boundary
layer(2000m). A fixedconcentration
is at the
upperboundaryadoptedfrom observations
at 2500 m for each
constituent
(25 ppbfor O3).
,
16
LOCAL TIME
Results of model runs are summarized in Table 3. The 24-hour
2-20 m
average
O3 flux to thecanopy
computed
using"standard"
model
parameters
[JacobandWofsy,
thisissue],1.07x10
•l molecules
10 .
cm-2 s-1, agreeswellwithobserved
values.Photochemical
productionand loss ratesfor ozoneare significantthroughoutthe
6 _
column[JacobandWofsy,thisissue]despitetheverylow levels
6
of NO; howeverthe net effectof photochemistry
is smalldueto
.
cancellation
of netproduction
atlowaltitudes
bynetlossathigher
levels. Net O3 loss,duemainlyto deposition
to the canopy,is
onlypartlybalanced
by supplyfwm thefreetroposphere
(fluxat
2500 m), leadingto decayof theozonecolumnwith a timeconstantof about2.5 weeks. We believethatozoneis replenished
ep-
0•
6
I
I
i_
6
I0
12
ß
I
I
14
16
isodicallyby deepmixingof thetroposphere
in convective
storms
[r:_•,,o,,,,g
.,, ,t ,,,;,,;..... • indicat;.ng
o" ;'""po
....... '•'"•'""deep
convection
in supplyingozoneto theloweratmosphere.
The factorsregulating
the O3 flux wereassessed
by imposing
arbitrarychanges
to thevaluesadopted
for leaf areadistributions
andfor ratesfor turbulenttransport(seeTable 3) in thismodel.
Increasing
theleafareaindexby a factorof 2 increased
thefluxto
thecanopy
by 12%,decreasing
by a factorof 2 decreased
theflux
by 16%. Results
weremoresensitive
therateof verticaltransport:
increasing
R, by a factorof 2 throughout
theboundarylayerdecreased
theozoneflux by 30%,decreasing
R, by 2 increased
flux
by 40%. Theresponse
to R, reflectschanges
in theconcentration
of ozonejustabovethecanopytop,i.e. changes
in thegradient
of
O3 in the "mixed"layer. We concludethattheflux of O3 to the
Amazonforestis significantly
limitedby therateof supplyfrom
the free troposphere,
althoughthe canopysurfaceresistance
also
has an effect. Note that the averageconcentration
of O3 in the
boundary
layeris regulated
largelyby theupperboundary
condition imposedin the model, implying an importantconnection
betweenO3 depositionfluxes and global-scaleprocesses
that
LOCAL TIME
20-30 m
10
_
8 _
$
E
J
0
œ
6
t
I
I
I
t
•
J
8
10
12
14
16
16
LOCAL TIME
determineozoneconcentrations
in themiddle troposphere.
6. CONCLUSIONS
Thenet ecosystem
exchange
of CO2in thetropicalforestun-
Fig.12.Observed
stomatal
resistances
forleaves
intheAmazon
forest
dergoes
a well-defined
diurnal
variation
driven
bytheinput
of
(range
given
byvertical
lines,
from
Roberts
etal.[1990])
and
values
calcu-solarradiation.
A curvilinear
relationship
wasfoundbetween
lated
using
the
model
ofBaldocchi
etal.[1987a,
hi.
solar
irradiance
anduptake
of CO2,withnetCO2uptake
ata
givensolarirradiance
equaltoratesobserved
overforests
in other
Daytime
aerodynamic
resistances
measured
between
27and41 climate
zones.
Uptake
during
thedaywasonaverage
closely
bal-
mwereoforder
0.1scm-1 (Trambore,
1988;S.M. FanandS.C. anced
byrelease
atnight,
withsoils
accounting
formorethan
90%
Wofsy,unpublished
data,1989],smaller
thanthecanopy
surfaceofnocturnal
respiration.
resistance
(Table
2),which
implies
thattheO3fluxisnotlimited Thecarbon
balance
ofthesystem
appeared
tobeverysensitive
byturbulent
transport
in andabove
thecanopy.
However,
the tocloud
cover
onthetimescale
oftheexperiment.
Theglobal
im-
strong
vertical
gradients
forO3observed
fromaircraft
[Gregory
et portante
ofcarbon
storage
inthetropics
haslong
been
recognized
ABLE-2B
FAN [:'i AL.: CARB½)N DI½}XiI_)E AN[) OZf)NE
EXf'i-iANGE
16,863
TABLE3. OzoneBudgets
overtheAmazonForest
Standard
[Not
3.4
LeafAreaIndex
3.3
Rax2* .
3.3
[O3]g
O3 flux at 2500 m
03 flux at canopytop
13.6
5.9(10)
9.9(10)
12.9
6.6(10)
1.10(11)
14.5
4.9(10)
8.4(10)
14.0
2.9(10)
6.9(10
Photochemical
production
#
-2.5(9)
-6.5(8)
-2.6(9)
-1.0(10)
oaz
R•x2õ
3.4
13.6
1.10(11)
1.4(11)
2.3(9)
-4.25(10)
4.46(10)-3.76(10)
-5.0(10)-2.77(10)
Resultsfrom simulationsusingthe modelof Jacoband Wofsy[1990], usingvaluesof canopyleaf areaindex
(LAD and aerodynamic
resistances
(Rs) from their standardmodelas well as higherandlower valuesof LAI.
All resultsshownas 24-houraverages.NO and03 concentrations
represent
meanvaluesbetween30 and2000
m, in pptv and ppbv respectively.Photochemical
productionis given as the columnintegralbetween30 and
2000m,in molecules
cm-2s
-1. Fluxes
andcolumn
timederivative
arealsogivenin molecules
cm-2s
-1. The
expression
5.6(10)denotes
5.6x1010.
* Withleafareadistribution
(2,4,8).
t Withleafareadistribution
(0.5,1,2).
• All aerodynamic
resistances
fromtheground
to2500m aremultiplied
ordivided
byafactor
of2.
õAveraged
between
30and2500m.
# Columnlossratebetween30 and2500 m.
[e.g.,Houghtonet al., 1985], andpresentresultssuggestthatglo- Baldocchi,
D.D, S.B.Verma,andD.E. Anderson,
Canopy
photosynthesis
hal carbonstoragemightbe affectedby changesin insolationas- andwater-use
efficiency
in a deciduous
forest,J. Appl.Ecol.,24, 251sociatedwith tropicalclimaticfluctuations
suchastheENSO.
260,1987.
Theforestwasanefficientsinkfor O3duringtheday. A large Browell,
E.V.,G.L.Gregory,
R.C.Hatriss,
andV.W.J.H.
Kircchoff,
Tro-
fluxwasobserved
(2.3'10
l• molecules
cm-2s
q) despite
lowam-
pospheric
ozone
andaerosol
distributions
across
theAmazon
basin,
J.
bient
concentrations
(- 6ppbv)
atcanopy
top.ThehighdepositionGeophys.
Res.,
93,1431-1451,
1988.
velocity
forO3during
theday(1.8cms-• atcanopy
top)appeared
Campbell,
G.S.,
AnIntroduction
toEnvironmental
Biophysics,
Heidelberg
toreflect
thelarge
stomatal
areaavailable
inthedense
foliage
of Science
Library,
1977.
thetrapical
f•est.
Denmead,
O.T.,and
E.F.Bradley,
Flux-gradient
relationships
inaforest
Rapid
deposition
ofO3over
theAmazon
forest,
combined
with canopy,
in
The
Forest-Atmosphere
Interaction
(edited
by
B.
A.
HutchinsonandB. B. Hicks,pp. 421-442, D. Reidel,Hingham,Mass.),1985.
the absenceof significantnet photochemical
productionandclose
Denmead,O. T., and E. F. Bradley,On scalartransportin plantcanopies,
couplingto the middle troposphere,suggestthat the Amazon Irrig. Sci.8, 131-149,1987.
forests
could
beasignificant
sink
forglobal
ozone
during
thewet Desjardins,
R.L.,
E.J.
Braek,
P.Alvo,
and
P.H.
Schuepp,
Aircraft
monitorperiod
(typically
9 months
oftheyear).Thissinkfor03isnon- ingofsurface
carbon
dioxide
exchange,
Science,
216,
733-735,
1982.
existent
during
thedryseason,
because
photochemical
production
Desjardins,
R.L.,J.L.MaePherson,
P.Alvo,
andP.H.Schuepp,
Measureof03 isenhanced
byhigher
levelsofNO,derived
formhigher
soil
ment
ofturbulent
heatandCO2 exchange
overforests
fromaircraft,
in:
emissions
andfrom biomassburning[Crutzenet al., 1985;Jacob TheForest-Atmosphere
Interactions,
ed.by B.A. Hutchinson
andB.B.
andWofsy,1988;Andreaeet al., 1988]. Deforestation
andcoloni- Hicks,D. Reidel,Hingham,
Mass.,pp645-658,
1985.
zationof theAmazonBasincoulddramatically
alterregional,and Droppo,
J.G.,Jr.,Concurrent
measurements
ofozone
drydeposition
using
perhaps
global,budgets
for 0 3, by simultaneously
decreasing
the
eddycorrelation
andprofilefluxmethods,
J. Geophys.
Res.,90,2112-
potential
for deposition
andincreasing
ratesfor photochemical2118,1985.
production.
Farquhar,
G.D.,S.yonCaemmerer,
andJ.A.Berry,
A biochemical
model
of photosynthetic
CO2 assimilationin leavesof C3 species,
Planta, 149,
Acknowledgements.
Wegratefully
acknowledge
ongoing
collaboration78-90,1980.
with
theABLE2B
science
team
and
with
themany
individuals
who
helped
Fi,'migan,
J.J.,Turbulent
transport
in flexible
plancanopies,
in The
in
the
execution
of
the
field
study
under
difficult
conditions,
particularly
Forest.Atmosphere
Interaction,
edited
by
B.
M.
Hutchinson
and
B.B.
JoseScolar(UNESP, Baum, S.P.) and OsvaldoM. R. Cabral (EMBRAHicks,pp. 443-480,D. Reidel,Hingham,Mass.,1985.
PA, Manaus,AM). This work was supported
by NASA grantsNAG1-55
Fitzjarrald,D.R., B.L Stormwind,G. Fisch,andO.M.R. Cabral,Turbulent
andNAGS-719,andby NSF grantATM-8413153to HarvardUniversity,
observed
just abovethe Amazonforest,J. Geophys.Res.,93,
and by NASA grant NAG-I-692 to the AtmosphericSciencesResearch transport
1551-1563,
1988.
Centerat theStateUniversityof New Yorkat Albany.
FitzjarraldD.R., K.E. Moore, O. Cabral,J. Scolar,A.O. Manzi, and L.C.
de AbreuSa, Daytimeturbulentexchange
betweenthe AmazonForest
andtheatmosphere,
J. Geophys.
Res.,thisissue.
Alvo,P.R.,R.C. Desjardins,
P.H.Schuepp,
andJ.I MacPherson,
Aircraft Fung,I., Analysis
of theseasonal
andgeographical
patterns
of atmospheric
measuremenu
of CO2 exchange
overvariousecosystems,
Boundary CO2distributions
witha three-dimensional
tracer
model,in TheChangLayerMeteorol.,
29, 167-183,1984.
ingCarbonCycle:A GlobalAnalysis,
editedbyJ.R.Trabalka
andD.E.
Bakwin,P.S.,S.C.Wofsy,andS.-M.Fan,Measurements
of reactive
nitro- Reichle,
pp.459-473,Springer-Verlag,
NewYork,1986.
genoxides
(NOy)withinandabovea tropical
forest
canopy,
in thewet Galbally,
I.E., andC.R.Roy,Destruction
of ozoneat theEarth's
surface,
season,
J. Geophys.
Res.,thisissue(a).
Q.J. R. Meteorol.Soc.,106,599-620,1980.
Bakwin,P.S.,S.C. Wofsy,S.-M. Fan,M. Keller,S. Trambore,andJ.M. Goreau,TJ., andW.Z. DeMello,Effectsof deforestation
on sources
and
daCosta,Emission
of nitricoxide(NO) fromtropicalforestsoilsandexsinksof atmospheric
carbondioxide,nitrousoxide,andmethane
from
changeof NO between
the forestcanopyandatmospheric
boundary someAmazonian
biotaandsoils,in Workshop
onBiogeochemistry
of
layers,
J. Geophys.
Res.,thisissue(b).
Tropical
RainForest:Problems
for Research,
editedbyD. Athaie,T.E.
Baldocehi,
D.D, B.B, Hicks,andP. Camaro,A canopystomatalresistance Lovejoy,and P. deM. Oyens,Universidadde Sao Pauloand World
modelfor gaseous
deposition
to vegetated
surfaces,
Atrnos.Environ.,21,
Wildlife Fund,Diracicaba,
S.P.,Brazil,1985.
91-101,1987.
Gregory,
G.L, C.H. Hudgins,
andR. A. Edahl,Jr.,Laboratory
evaluation
16,864
FAN E'I •,L.: CARBON DIt)XIDE
AND OZ()NE
EX('t!ANGE
ABLE-2B
of an airborne ozone instrument which compensates
for
mation,U.S.AtomicEnergy
Commission,
Washington,
D.C., 1970.
altitude/sensitivity
effects,
Environ.Sci.Technol.,
17, 100-103,1983.
Odum,H.T., G. DrewryandJ.R.Kline,ClimateatE1Verde,1963-1966,
in
Gregory,
G.L., E.V. Browell,andL,S. Warren,Boundary
layerozone:a
A tropicalRain Forest,editedby H.T. OdumandR.F. Pigeon,pp.
airborne
surveyabovethe Amazonbasin,J. Geophys.
Res.,93, 1452- BB47-B418,
Divisionof Technical
Information,
U.S. AtomicEnergy
1468,1988.
Commission,
Washington,
D.C., 1970.
Gregory,G.L., E. Browell,L. Warren,and C. Hudgins,AmazonBasin Ohtaki,E., Turbulent
transport
of carbon
dio•ddeovera paddyfield,Bounozoneandaerosol:Wet season
observations,
J. Geophys.
Res.,thisisdaryLayerMeteorol.,19, 315-336,1980.
sue.
Ratisbona,L.R., The climateof Brazil, in World Surveyof Climatology,
Hicks,B. B., andR. T. Mc3/Iillan,On themeasurement
of drydeposition vol. 12,editedby W. Schwerdfferger,
Elsevier,NewYork, 1976.
using impeffectsensorsand in non-idealterrain,BoundaryLayer Roberts,
J., O. M. R. Cabral,andL. F. deAguiar,StomatalandboundaryMeteorol.42, 79-94, 1988.
layerconductances
measured
in a TerraFirmerainforest,J. Appl.Ecol.,
Hicks,B.B.,D.D. Baldoechi,
T.P.Meyers,R.P.Hosker,andD.R.Matt,A
in press,1990.
preliminary
multipleresistance
routinefor derivingdry deposition
velo- Schlesinger,
W.H., Carbonbalance
in terrestrial
detritus,
Annu.Rev. Ecol.
citiesfrom measuredquantities,Water Air Soil Pollut., 36, 311-330, Syst.,8,51-81,1977.
1987.
Sehmel,G. A., Particleandgasdry deposition:A review,Atmos.EnHoughton,R. A., R. D. Boone, J. M. Melillo, C. A. Palm, G.M.
viron.,14, 983-1011,1980.
Woodwell,
N. Myers,B. Moore1ll,andD. L. Skole,Netfluxof carbonShuttleworth,
W.J.,J.H.C.Gash,C.R.Lloyd,C.J.Moore,J.Roberts,
A.O.
dioxide
fromtropical
forests
in 1980,Nature316,617-620,1985.
Marques,
F. Fisch,V.P. deSilva,M.N.G.Ribeiro,L.C.B.Molion,L.D.
Jacob,
D.J.,andS.C.Wofsy,Photochemistry
of biogenic
Emissions
over AbreuSa,J.C.A.Nobre,O.M,R.Cabral,
S.R.Patel,andJ.C.deMoraes,
theAmazon
forest,
J.Geophys.
Res.,
93,1477-1486,
1988.
Eddycorrelation
measurements
of energy
partition
for Amazonian
Jacob,
D.J.,andS.C.Wofsy,Budgets
of reactive
nitrogen,
hydrocarbons,
forest,
Q.J.R.Meteorol.
Soc.,110,1143-1162,
1984a.
andozone
overtheAmazon
forest
during
thewetseason,
J. Geophys.
Shuttleworth,
W.J.,J.H.C.Gash,
C.R.Lloyd,
C.J.Moore,
J.Roberts,
A.O.
Res.,
thisissue.
Marques,
F. Fisch,
V.P.deSilva,M.N.G.Ribeiro,
L.C.B.Motion,
L.D.
Kaplan,
W.A.,S.C.Wofsy,
M. Keller,
andJ.M.daCosta,,Emission
ofNO Abreu
Sa,J.C.A.Nobre,
O.M.R.Cabral,
S.R.Patel,
andJ.C.deMoraes,
anddeposition
of O3in a tropical
forest
system,
J. Geophys.
Res.,95, Observations
of radiation
exchange
above
andbelow
Amazonian
forest,
1389-1395,
1988.
Q.J.R.Meteorol.
Soc.,
110,1163-1169,
1984b.
Keller,
M., W.A.Kaplan,
andS.C.Wofsy,
Emissions
ofN20,CH4,and Trambore,
S.E.,Carbon
cycling
andgasexchange
in softs.Ph.D.thesis,
COg from tropicalforestsofts,J. Geophys.
Res.,91, 11,791-11,802, 194 pp., ColumbiaUniv., New York, 1988.
1986.
Verma,S.B., andN.J. Rosenberg,
Carbondioxideconcentration
andflux
Kirchhoff,
V.WJ.H.,E.V. Browell,andG.L. Gregory,
Ozonemeasure- in a largeagriculturalregionof greatplainsof North America,J. Geomentsin the troposphere
of an Amazonrainforestenvironment,
J. Geo-
phys.Res.,81,399-405, 1976.
phys.Res.,93, 15,850-15,860,
1988.
Verstraete,M.M., Radiationtransferin plant canopies:
Transmission
of
Kirchhoff,V.W.J.H.,i.M.O. da Silva,andE.V. Browell,Ozonemeasure- directsolarradiationandtheroleof leaf orientation,
J. Geophys.
Res.,
meatsin Amazonia:Dry season
versuswet season,
J. Geophys.
Res., 92, 10,985-10,995, I987.
this issue.
Webb,E.K., G.I. Pearman,andR. Leuning,Correctionof flux measureLemon,E.H., L.H. Allen,andL. Miller, Carbondioxideexchange
of a
meatsfor densityeffectsdue to heat and watervapourtransfer,Q. J.
tropicalrainforest,1I,Bioscience,
20, 1054-1059,1970.
Roy. Meteorol.Sot., 106, 85-100, 1980.
Lenschow,
D.H.,andL. Kristensen,
Uncorrelated
noisein turbulence
Wesely,
M.L.,Improved
parameterizations
forsurface
resistance
to gasemeasuremeats,
J.Atmos.
Ocean.
Technol.,
2, 68-81,1985.
ousdrydeposition
in regional
scale,
numerical
models,
EPAAgreement
Len•chow,
D.H.,R. Pearson,
Jr.,andB.B.Stankov,
Estimating
theozone DW89932394-01,
Environ.
Prot.Agency,
Washington,
D.C., 1989.
budget
in theboundary
layerbyuseof aircraft
measurements
of ozoneWesely,
M.L.,andR.L.Hart,Variability
ofshort
termeddy-correlation
eseddy,
fluxandmeanconcentration,
J. Geophys.
Res.,86,7291-7292, timares
ofmass
exchange,
inTheForest-Atmosphere
Interaction,
edited
1981.
byB.H.Hutchinson
andB.B.Hicks,
pp.591-612,
D. Reidel,
Hingham,
Lenschow,
D.H., R. Pearson,
Jr., and B.B. Stankov,Measurements
of
Mass.,1985.
'ozone
vertical
fluxtoocean
andforest,
J. Geophys.
Res.,87,8833-8837,Wesely,
M.L.,J.A.Eastman,
D.R.Cook,andB.B.Hicks,Daytime
varia1982.
tionsof ozoneeddyfluxesto maize,BoundaryLayerMeteorol.,15,
Liu,S.C.,Modelstudies
of background
ozoneformation,
in Tropospheric 361-373,1978.
Ozone,
Regional
andGlobalSealInteractions,
NATOASISeries,
edited Wesely,
M.L., D.R.Cook,andE.M.Williams,
Fieldmeasurement
of small
byI.S.A.'Isaksen,
D.Reidel,
Hingham,
Mass.,1988.
ozonefluxes to snow,wet bare soil, and lake water,BoundaryLayer
Lumley,J.L.,andH.A. Panofsky,
TheStructure
of Atmospheric
Tur- Meteorol., 20, 459-471, 1981.
bulence,
JohnWiley, New York, 1964.
Wofsy,S.C.,R.C. Harriss,andW.A. Kaplan,Carbondioxidein the atmo-
Matthews,
E., Globalvegetation
andlanduse:Newhighresolution
data sphereovertheAmazonbasin,J. Geophys:
Res.,93, 1377-1387,1988.
bases
forclimate
studies,
J. Appl.Meteorol.,
22,474-487,1983.
Zimmerman,P.R., J.P. Greenberg,
andC.E. Westberg,Measurements
of
MclVIiilen,
R.T., Aneddycorrelation
technique
withextended
applicabili- atmospheric
hydrocarbons
andbiogenicemissionfluxesin the Amazon
ty tonon-simple
terrain,Boundary
LayerMeteorol.,
43, 231-245,1988.
boundary
layer,J. Geophys.
Res.,93, 1407-1416,1988.
Medina,
E., andH. Klinge,Productivity
of tropical
forests
andtropical
woodlands,
Encycl.PlantPhysiology,
12D,281-304,1983.
Meyers,
T.P.,
and
D.D.
Baldocchi,
Acomparison
ofmodels
forderiving
s.-M.
Fan,
S.C.
Wofsy,
P.S.
Bakwin,
and
D.J.Jacob
Division
ofApplied
drydeposition
fluxes
ofO3andSO2toa forest
canopy,
Tellus,
40B, Sciences
andDepartment
ofEarth
andPlanetary
Sciences,
Harvard
Univer270-284,1988.
sity,Cambridge,
MA 02138.
Mooney,
H.A.,P.M.Vitousek,
andP.A.Matson,
Exchange
ofmaterials
D.R.Fitzjarrald,
Atmopsheric
Sciences
Research
Center,
State
University
between
terrestrial
ecosystems
andtheatmosphere,
Science,
238,926- ofNewYork
atAlbany,
Albany,
NY 12205.
932, 1987.
Odum,
H.T.,andC.F.Jordan,
Metabolism
andevapotranspiration
ofthe
(Received
June
8,1989;
lowerforestin a giantcylinder,in A TropicalRainForest,editedby
revised
February
20, 1990;
H.T.OdumandR.F.Pigeon,
pp.1165-1189,
Division
ofTechnical
Infor-
accepted
March
5, 1990.)