UBC Faculty Research and Publications

Spatial and temporal variability of CO2 concentration and flux in a boreal aspen forest Yang, P. C.; Black, T. Andrew; Neumann, Herman H.; Novak, M. D.; Blanken, P. D. Nov 30, 1999

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

Item Metadata


52383-Blacki_AGU_1999JD900295.pdf [ 736.36kB ]
JSON: 52383-1.0041907.json
JSON-LD: 52383-1.0041907-ld.json
RDF/XML (Pretty): 52383-1.0041907-rdf.xml
RDF/JSON: 52383-1.0041907-rdf.json
Turtle: 52383-1.0041907-turtle.txt
N-Triples: 52383-1.0041907-rdf-ntriples.txt
Original Record: 52383-1.0041907-source.json
Full Text

Full Text

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 104, NO. D22, PAGES 27,653-27,661, NOVEMBER 27, 1999  Spatial and temporal variability of COz concentrationand flux in a boreal aspenforest P.C.Yang, • T. A. Black, • H. H. Neumann, 2M.D. Novak, • andP.D. Blanken 3 Abstract. In conjunction witheddycovariance measurements of CO2fluxesat the39.5-mheight over a 21.5-m-tallborealaspenstandin northernSaskatchewan, CO2 concentration wasmeasuredat eightheightsin orderto calculatenet ecosystem exchange.Duringbothleaflessandfull-leaf peri-  ods,daytime vertical CO2concentration gradients above 9 rnwereweak(< 0.2gmolmo1-1 m't),but were strongbelowthisheight. Little changein CO2storagein the air columnbelow 39.5 rn occurredduringmuchof thedaytime,while aroundsunriseandsunsetCO2storagechangedmainly below9 m. For the restof the night,over85% of the increasein CO2storageoccurredabove9 m. On somecalmnightsduringthe growingseason,CO2alsoaccumulated below9 rn resultingin a suddenupwardCO2flux at 39.5 rn followingtheresumptionof mixing2-3 hoursaftersunrise.A 1O-dayexperimentwasconducted to determinethespatialvariabilityof CO2flux in thetrunkspace. Two eddycovariance systems weremountedjust abovetheunderstory abouttwo treeheightsapart.  Thecorrelation between CO2fluxeswerepoorevenunderunstable (daytime) conditions, suggesting  arelatively heterogeneous understory and soil.Incontrast, thecorrelation between water vapor  fluxeswashigh(r = 0.70) in unstable conditions.However,averagedaytimeandnighttimeCO2 fluxesoverthe 10 daysagreedto within5%. Thissuggests thatpartitioningnetecosystem exchange betweenoverstoryandunderstory on an hourlybasisusinga single-understory eddycovariance systemis inadvisable; however,partitioning probablycanbe donequitereliablyusing5-dayaverage fluxes.  1. Introduction  trunkspaceare similarto thoseabovethe forestat the Old Aspen (OA) sitein thisstudy[Blankenet al., 1998]. The eddycovariancetechniquehasbeenwidely and successOne of the importantusesof the CO2 flux measuredusingthe fully usedin studiesof CO2exchange betweenvariousforestsand eddycovariancetechnique(eddyCO2flux F½)is to estimatethe theatmosphere [e.g.,Vermaet al., 1986; Wofsyet al., 1993;Hol- net ecosystem exchange(NEE) betweenthe atmosphere andthe linger et al., 1994, 1998; Grace et al., 1995; Fan et al., 1995; soil or vegetation[e.g., Wofsyet al., 1993; Black et al., 1996' Baldocchiand Vogel, 1996; Black et al., 1996; Jarvis et al., Grecoand Baldocchi,1996; Jarvis et al., 1997; Hollinger et al., 1997]. Someresearchers have also successfully usedthis tech- 1998]. This exchange,which is often referredto as the biotic niqueto studyCO2exchange abovetheforestfloor(e.g.,Baldoc- flux, can be expressedas NEE = F½+ ASa/At, where Z•a/At is the chiandMeyers[1991]in anoak-hickory forestin Tennessee, Lee changeof the CO2storagein the air columnbeneaththe eddyflux et al. [1994] in a hemlock-Douglas-fir forestin coastalBritish sensors.The estimationof ASa/At requireshalf-hourlymeasureColumbia,Black et al. [1996] in a borealaspenforest,andBalmentsof CO2 concentration at variousheightsbetweenthe eddy docchiet al. [1997]in a borealjack pineforest). The relatively covarianceinstruments and the ground(referredto as the CO2 small contributionsof small eddiesto turbulenttransportabove concentration profile). Severalquestions needto be answered in forestsmake it feasibleto use the eddy covariancetechniqueto carryingOutthisprocedure.First,how doesthe distribution of measureCO2 fluxes above the forest, even using a closed-path CO2 beneaththe eddycovariancesensorsdependon time of day  infraredgasanalyzer(IRGA) systemwith a relativelylongsam- and atmosphericconditions?Second,how many samplinglevels plingtube[Leeet al., 1999]. Large,intermittent eddystructures in the profile systemare needed to properly estimate •a/•t?  that penetratethrough the overstoryenable eddy covariance Third, how importantare the half-hourlychangesof CO2 storage measurementsto be made near the forest floor [Baldocchi and in the air column in determiningthe pattern of NEE during the Meyers,1991;Blankenet al., 1998]. Furthermore, cospectral day? The answersto thesequestionsare importantfor obtaining analysishas shown that turbulenttransfermechanismsin the reliable diurnal NEE patterns at forest CO2 flux measurement  1Faculty ofAgricultural Sciences, University ofBritish Columbia, Vancouver, Canada.  2Atmospheric Environment Service, Downsview, Ontario, Canada. 3Department ofGeography andEnvironmental Studies, University of Colorado, Boulder.  Copyright1999 by the AmericanGeophysicalUnion.  sites. A relatedquestionis how spatiallyvariableis the CO2 flux abovethe understory?The answerto this questionwill help determinethe feasibilityof subtractingCO2 fluxes measuredat one or more locationsin the trunk spacefrom thoseabove the forest  to provideanestimate of •heCO2source or sinkstrength of the overstory. The flux measuredat a singlelocationabovethe understoryis representativeof a much smaller area than that of a flux measuredabovethe forest. For example,Blanken[1997],  usingthe footprintmodeldeveloped by Schueppet al. [1990],  Papernumber1999JD900295.  found that at the OA site the upwind distancewhere the vegeta-  0148-0227/99/1999JD900295509.00  tion makes the maximum  27,653  contribution  to the measured flux at the  27,654  YANG ET AL.' VARIABILITY OF COz FLUX IN AN ASPEN FOREST  39.5-m heightwas 100 and 300 m for typicaldaytimeand nighttime conditions,respectively.In contrast,the corresponding val-  Half-hourly CO2 fluxes were measuredusing the eddy covariance method at the 39.5-m height above the ground on a 37-m  uesfor the4-m height(in thetrunkspace)wereabout10 and40 walk-upscaffoldtower (maintower)in 1993, 1994, and 1996. m. Usinga Lagrangian randomwalk model,Baldocchi[1997] Fluctuationsin the wind vectorcomponentswere measuredusing found that flux footprintsbeneaththe overstorywere even more contractedwhen horizontal wind velocity fluctuationswere considered. This meansthat the validity of subtractingthe two fluxes  is questionable whenthereis markedhorizontalvariabilityof the vertical fluxes in the trunk space. One way of assessingthis is to determinethe horizontalvariability of eddy fluxes in the trunk  a three-dimensional  sonic  anemometer-thermometer:  a model  DAT-310 with model TR-61B probe, Kaijo-Denki Co., Tokyo,  Japan (20-cm path length) in 1993 and 1994 and a model 1012R2A, Gill Instruments, Lymington,England(15-cm path length)in 1996. Fluctuations in CO2andwatervaporconcentra-  tion were measuredusing the closed-pathmethod with a temperature-controlled LI-COR 6262 IRGA. Air wasdrawnthrough Thereforethe specificobjectives of thispaperare (1) to learn a heatedtube (6 m long by 3.2 mm ID Bev-a-linein 1993 and how the vertical distributionof CO2 below the eddy covariance 1994 and 4.7 m long by 4 mm ID Dekoron in 1996) at space.  sensors depends on timeof day andatmospheric conditions, (2)  6.5L min-• in 1993and1994andat 10L min'• in 1996insuring  to determinehow the accuracyof the estimationof ASa/Atde- turbulent flow. Using a similar eddy covariancesystem,CO2 pendson the numberof samplingheightsin the concentration fluxesabovethe understorywere measuredcontinuouslyat the 4-  profilesystem,(3) to assess the relativeimportance of ASa/Atin theprocessof CO2exchangebetweenthe atmosphere andtheforest,and (4) to assess the horizontalvariabilityof COz flux in the lowertrunkspaceabovetheunderstory.  2. Experimental Methods 2.1 Site Description The OA forest site (53.63øN, 106.20øW) is located in the  southernpartof PrinceAlbertNationalPark,Saskatchewan, Can-  ada. The siteis an even-aged standof tremblingaspen(Populus tremuloides Michx.) with scattered balsampoplar(PopulusbalsamiferaL.). In 1994, the standwas70 yearsold, meancanopy heightwas 21.5 m, diameterat the 1.3-m heightwas 20 cm, and stemdensitywas 830 stemsper hectare. The trunk spacewith almostno branchesextendedup to about15 m. The understory  was mainlycomposed of 2-m-tall hazelnut(Coryluscornuta Marsh.) with occasional clumpsof alder (Alnuscrispa (Ait.) Pursch)and sparseshrubs(e.g., prickly rose,Rosa acicularis  m heighton a 6-m scaffoldtower40 m (abouttwo treeheights) awayfrom the main tower in 1993 and 1994. Correctionswere madefor the effectof fluctuationsin air densityon the CO2fluxes [Webb et al., 1980]. A two-dimensionalcoordinaterotation was applied to make the averagevertical and lateral wind velocity  components equalto zeroabovetheforest[Tannerand Thurtell, 1969], and a one-dimensional rotationwas appliedto bring the lateral velocity component to zero above the understory [Baldocchiand Hutchison,1987]. A detaileddescriptionof the eddy covariancesystemsand analyticalprocedurescan be found in the work of Black et al. [1996] and Chen et al. [1999]. Comparisonsof the three eddy covariancesystems,with the sensors mountedabove the forest,made at Camp Borden,Ontario, Canada [Leeet al., 1996] and at the site [Yang,1998] indicatedthat CO2fluxesagreedto within 7%. During the 1O-dayperiod,August12-22, 1994, an experiment wasconductedto determinethe degreeof spatialvariabilityin the CO2fluxesabovethe hazelnutunderstory.Eddy fluxesof CO2 at the 5.9-m height on the main tower were made using a KaijoDenki DAT-310 sonic anemometeralready operating at this height and an IRGA identicalto that on the 6-m tower. These fluxes were comparedwith those being measuredat the 4-m heighton the 6-m tower. Becausethe groundlevel at the main  Lindl.). The soil, an Orthic Gray Luvisol,has a surfaceorganic layerabout8-10 cmthickabovea silty-claytexturedsubsoil.The topography is relativelylevel,andthefetchis at least3 km in all directions. More detailedsite descriptionscan be found in the tower was 0.3-0.5 m lower than that at the 6-m tower, the elevation difference between the two measurementheights was only work of Black et al. [1996] and Chenet al. [1999]. 1.3-1.6 m, whichprobablyhad little effecton the CO2flux com2.2 COz Flux and Concentration-Profile Measurements parison.To compareCO2concentrations at the 4-m heightat the two towers, CO2 concentrationson the main tower were calcuDuring late fall of 1993 and much of 1994, COz concentralated using a rectangularhyperbolicfit to the concentrations tions were measured at 0.8, 2.3, 9.5, 15.7, 18.8, 21.9, 25.0, and measuredat the eight heights. 34.2 m abovethe ground. With an eight-inputmanifold,a 200 L  rain-• rotarypumpdrewairdowna 9.3-raminnerdiameter (ID) Dekoron tubefromeachheight(25L min'• perheight) to a data logginghut. The tubeswere heatedto preventcondensation using a 22-gaugebare nichromewire passinginside. Using eight solenoidvalvesin the hut, air from each level was sequentially pumpedusinga diaphragmpumpthroughan IRGA (model6262, LI-COR Inc., Lincoln,Nebraska).A similarsystem,butwith unheatedtubes,wasusedduringJuly throughOctober1996 [Chen et al., 1999]. COz concentrations at all eightlevelsweremeasuredfor approximately1 min twice everyhalf hour, and the system wascalibratedautomaticallyevery6 hours. Changesin COz storagein the air column beneaththe eddy covariancesensors were calculatedfrom theseCO2 concentrationprofile data. The rate of changefor a given half hour was estimatedby calculating the differencebetweenthe meanair COz storagein the previous andfollowinghalf hours.  3. Results  and Discussion  3.1 Diurnal Courses of CO2 Concentration and the Estimation of ASa/At  Figure1 showsthe ensemble-averaged diurnalcoursesof CO2 concentration duringthe leaflessand full-leaf periodsat the OA sitein 1994. NighttimeanddaytimeCO2concentrations from the 9.5-m heightto the 39.5-m heightwerevery similar(magnitudes  ofvertical concentration gradients <0.2gmolmol'• m'l),showing thatconsiderable turbulentmixingextendedbelow the bottomof the aspencanopy. Mixing within aspencanopyon calm nights wasclearlyindicatedby reducedverticalair temperature and CO2 concentration gradientsbetweenthe 15- and21-m heights,which oftenextendeddownto 9 m. This mixingwaslikely due to cold  YANG ET AL.' VARIABILITY OF COz FLUX IN AN ASPEN FOREST  27,655  (a)  37O  366 362 'E  8 5ooõ  (b)  O0400 0  4  8  12  16  20  24  Local Time (CST)  Figure1. Ensemble-averaged CO2concentrations (mole fraction ingmolmol'l moist air)ateightheights (0.8,2.3, 9.5,15.7,18.8,21.9,25.0,and34.2m)attheOldAspensitein 1994for(a)theleafless (February 4 toApril10)periodand(b) thefull-leaf(June1 to August31) period.The CO2concentration at an additional height(0.5 m, circles)duringthe summerwasmeasured usinga differentgasanalyzer.The symbolsare squares(0.8 m), inverted triangles (2.3 m), triangles (9.5 m), leftpointingtriangles (15.7m), rightpointingtriangles (18.8m), pluses(21.9 m), asterisks(25.0 m), andcrosses(34.2 m).  air shedding from foliageas a resultof radiativecooling[Yang, 1998]. NighttimeCO2 concentrations below 9 m were signifi-  the densityof the air in layer i, and Ci is the CO2 concentration  cantly higherthan thoseabove due to the strongtemperatureinversion in this layer, which suppressedthe upward turbulent transportof CO2. This inversionwas the resultof the radiative coolingof the hazelnutunderstorydue to the opennessof the aspen canopy. Concentrationsbelow 9 m changedmost rapidly near sunriseand sunset,and during the full-leaf period the mag-  exceptfor the bottomand top layers,was approximatedby halving the differencebetweenthe samplingheightaboveand below.  nitude ofthese changes was20-50gmolmol'l h'l (Figure lb). In  (ILl. molmo1-1 wetair)in layeri. Thethickness of eachlayer(Azi), Thecoefficients ofdetermination (r2)resulting fromthelinearregressionof changesin storageusingall heightson the valuescalculatedusingall samplingheightcombinationsare plottedagainst numberof samplingheightsusedduringthe leaflessand full-leaf  periods(Figure2). The plotsclearlyshowthatoneheightis in-  the eveningthe rapid increasein concentrationusuallyceasedat  adequateto estimateAS``/Atregardless of whichheightis chosen. about22:00 LT (CST), and for the restof the night,concentra- Furthermore,the plots show three relatively distinct groupsof tionsbelow 9 m remainedquite constant. In contrast,concentra- data:(1) thereis at leastone heightaboveand eithertwo heights tions betweenthe 9- and 39.5-m heightsincreasedonly slightly or onlythe2.3-mheightbelowthebaseof theaspencanopy(15duringthe 5 hoursprior to 22:00 LT but increasedsteadilyat 2-3 m height),(2) thereis no heightbelow9.5 m, and (3) the only gmolmol'l h'l during therestof thenight.Thecommon occur- heightbelow 15 m is the 0.8-m height. The first grouphasthe renceat the OA site of relativelyconstantCO2 concentrationbe- highest r2 values, andof thesethehighest alwaysincludes the low 9 m after 22:00 LT indicatesa quasi steadystate condition 2.3-m heightfor any numberof heightsfor leaflessand full-leaf with the rate of upward turbulentdiffusionthroughthe aspen periods. The latter is also shownin Table 1, which lists the canopybeing approximatelyequal to the effiux of respiratory maximum r2 valuefor eachnumber of sampling heights.The CO2fromthesoil. This is likelybecause(1) steadynighttimesoil other two groupsin Figure2 cannotbe usedto adequatelyestitemperature resultedin a relativelyconstant CO2effiuxand (2) mateAS``/At no matterhow many heightsare used. More effecthe friction velocity u, at the 39.5-m height during the period tive mixing within the standprior to leaf emergenceprobably 22:00-05:00LT wasalsorelativelyconstant(e.g.,for the full-leaf accounts forthehigher r2 values forthefirstandsecond groups periodin 1994 the ensembleaveragefor this periodvariedfrom duringthe leaflessperiod. TheseresultssuggestthatAS``/Atcan 0.27to 0.29m be reasonablyestimatedby usingat leastone height aboveand To determinehow many CO2 concentrationsamplingheights eithertwo heightsor only the 2.3-m heightbelowthe baseof the aspencanopy. If the choicewere restrictedto two heights,the are necessaryand to determinethe best choicesfor this site, the  changein CO2storagein the 0- to 39.5-m air column  best combination would be 2.3 and 15.7 m with a standard error  was calculatedusing all combinationsof the eight heightstaken 1, 2, 3, 4, 5, 6, 7, or 8 at a time. AS,,/At was calculatedusing  ofestimate (syx) of0.87lu, molm'2s'• during thefull-leaf period (Table 1). If the choicewere threeheights,the bestcombination  would be2.3,9.5,and25.0mwithansyx of0.58gmolm'2s'l. In thiscase,thedifference in r2 valuescaused by usinga different i:I  where M isthemolecular weight ofmoist air(: 29g mol'l),Piis  above-canopy height,namely,21.9 or 34.2 m, was only about 2%.  27,656  YANG ET AL.: VARIABILITY OF CO2FLUX IN AN ASPEN FOREST 1  0.8  0.6  0.4-  (a) leafless  6 1I  I 3I  I  2  4  I  5  I  6  (b) full-leaf  I  7  0  I 2I 3I  1  I  4  I  5  I  6  I  7  8  Numberof samplingheights  Figure2. Coefficients ofdetermination (rz)fromthelinear regression ofCOzstorage change inthe0-to39.5-m air columncalculated usingall eightsamplingheightson valuescalculated usingall othercombinations of heightsversusnumberof samplingheightsfor the sameperiodsas thosein Figure 1. Circles are combinationswith at least one heightaboveand eithertwo heightsor only the 2.3-m heightbelow the baseof the aspencanopy(15 m), crossesarethosewith no heightsbelow9.5 m, andsquaresarethosewith 0.8 m beingthe only heightbelow 15 m. The solid line connectsthe meansof combinationsin the first group,the dottedline connectsthe meansin the secondgroup,andthedash-dotted line connects the meansin thethirdgroup. 3.2 Diurnal Changesin AS./At and Its Effect on Net EcosystemExchange  NEE. Duringthe earlyeveningwith a sharpdropin photosynthesis,therewasa markedincrease in AS,/At asrespiration con-  Figure 3 showsthe ensemble-averaged diurnal patternof &S,/At for the 0- to 39.5-m air column,the eddyCOz flux Fc at the39.5-mheight,andNEE for thefull-leafperiodin 1994at the OA site. During this period,sunriseoccurredbetween5:00 and  tinuedwhile air and soil temperatures were high. The large increasein COz concentration in the air beneaththe 9-m heightdescribedin section3.1 accountedfor 65-90% of AS,/At between 17:00 and 22:00 LT. After 22:00 LT, the upwardeddy flux of COz as well as the rate of COz accumulationbecamerelatively  6:00 LT, and sunset occurred between 20:30 and 22:00 LT.  constant, averaging 3-4 and1 gmolm'z s'•, respectively. Over  Be-  tween 8:00 and 9:00 LT, AS,/At accountedfor 50-60% of the 85% of the latteroccurredin the layerbetweenthe 9- and 39.5-m NEE. During this hour and the followinghour, 65 and 60%, re- heightsas a resultof the relativelyunchangingCOz concentration spectively,of AS,/At wasdueto the decreasein the COz concen- beneaththe 9-m heightas discussed in section3.1. Similar pattrationof the air beneaththe 9-m height. Betweenlate morning ternsof the daily coursesof thesefluxeshavebeenobservedby and late afternoon,AS,/At made a negligiblecontribution.to otherworkers(e.g.,Fan et al. [1995]in a boreallichenwoodland  Table1. Statistics of theLinearRegression between COzStorage Change Calculated UsingAll Eight Sampling Heights(AS,/At)andCalculated Storage ChangeUsingtheBestCombination for a GivenNumber  (n)ofSampling Heights (AS,/At)' Numberof Heights,  rz  Syx,  BestCombination of  pmolm'zs'•  Sampling Heights, m  n  a  1  0.881  0.822  0.144  9.5  2 3 4 5 6 7  0.910 0.993 0.974 0.997 0.998 1.000  0.948 0.969 0.994 0.998 1.000 1.000  0.078 0.060 0.026 0.014 0.006 0.004  2.3, 15.7 2.3, 9.5, 25.0 0.8, 2.3, 9.5, 25.0 0.8, 2.3, 9.5, 18.8, 34.2 0.8, 2.3, 9.5, 15.7, 25.0, 34.2 0.8, 2.3, 9.5, 15.7, 18.8, 25.0, 34.2  1  0.592  0.599  1.862  9.5  2 3 4 5 6 7  0.766 0.947 0.950 0.986 0.994 0.999  0.913 0.962 0.988 0.995 0.999 1.000  0.866 0.576 0.326 0.205 0.085 0.048  2.3, 15.7 2.3, 9.5, 25.0 0.8, 2.3, 9.5, 25.0 0.8, 2.3, 9.5, 18.8, 34.2 0.8, 2.3, 9.5, 15.7, 25.0, 34.2 0.8, 2.3, 9.5, 15.7, 18.8, 25.0, 34.2  Leafless  Full-Leaf  Regression isAS./At = a(AS./At)' + b. Theabsolute values ofb areless than 0.004•tmol m-zS'1.$yx isthestandard error ofestimate. Allvalues ofcoefficient ofdetermination rzareshown inFigure 2.  YANG ET AL.' VARIABILITYOF CO2FLUX IN AN ASPENFOREST I  I  I  I  I  5 ,, .  /":••  --5  i  .." -10  27,657  _  .-",  ' !.................. _ I  -200  .  I  4  I  8  12  I  16  -  I  20  24  Local Time (CST)  Figure 3. Ensemble-averaged COz flux Fc measured usingthe eddycovariance systemat the 39.5-mheight (squares), the rateof changein COzstorage (Z•a/Z•) in the air columnbeneath the eddycovariance system (circles),andnetecosystem exchange (NEE) (Fc+ Z•a/At, dash-dotted line)duringthefull-leafperiodin 1994 at theOld Aspen(OA) site. Theverticalbarsindicateonestandard deviation. andJarviset al. [1997] in a borealblackspruceforest). In the theforestweresmallerthanusual. An exampleof this (July16latter study,bSa/At accountedfor about20% of the nighttime 17, 1996) is shownin Figure4. In this case,COz concentration eddyflux measuredat the 26-m height,which is slightlylessthan below the 9-m heightcontinuedto increasesomewhatsporadi-  callyuntilwell aftersunrise(8:00LT). At the0.5-mheight,COz reached 860 gmolmol'l around sunrise.Around [1997] alsofounda significant proportion of bSa/bt accumulated concentration  the corresponding fraction(35%) in this study. Jarviset al.  above and in the upper part of the canopy;however,CO•. also tendedto increasebelow the 6-m height. On very calm nights,CO•. continuedto accumulateafter midnight in the air beneaththe 9-m height while eddy fluxes above 55O  I  9:00 LT, the stableboundarylayer brokeup, resultingin a strong  mixingof CO2throughout the39.5-maircolumn(clearlyseenin Figure 4a).  At this time, there was a strongupward flux  (approximately 26gmolm'2s-1)atthe39.5-m height (Figure 4b).  I  I  I  I  I  I  I  I  I  I'  (a)  5OO  450 F 400' 350  •,  E  20-  0  -  (b)  j •r•  -  E  •  0  '•  .....'._................... •...... .,.,., ,...,,, ,...,', ...... 4,•,•.._.•.  x  =  -2:0 '  0 r,.)  18  I  I  I  I  I  22  2  6  10  14  18  Local Time (CST)  Figure 4. (a)Diurnal courses oftheaverage COz concentration (gmol mol 'l moist air)oftheairbetween the0-and 9-mheights (circles), the9- and20-mheights (squares), the20-and30-mheights (inverse triangles), andthe30and39.5-mheights (triangles) onJuly16-17,1996,attheOAsite.(b)Thecorresponding courses ofFcatthe39.5-  m height(solidline)andthe2-hourrunning meanof Z•a/At (dash-dotted line). ThestrongupwardCOzflux at 9:00-9:30LT (arrowin Figure4b) corresponds with the uniformhigh COz concentration as a resultof mixing (arrowin Figure4a).  27,658  YANG ET AL.: VARIABILITY OF CO2FLUX IN AN ASPEN FOREST 20  •  •  I  • o o  (a) unstable & neutral 10  oo  • 1:1 line o  o  -10  r 2 = 0.22  -20  vv  (b) stable  1'1 line  v  10  0  v  v  -10  -2-020 -15  ,• Iv  -10  I  -5  I  0  v  vv  v  v  r 2 = 0.08 I  5  I  10  I  15  6-m scaffoldtower(gmolm'2 S'1) Figure5. Comparison of thehalf-hourly valuesof nethazelnut understory CO2exchange (NHE) measured using eddycovariance systems atthe6-mscaffold andmaintowers(40m apart)in thetrunkspace duringAugust12-22, 1994,in (a) unstable andneutralconditions and(b) stableconditions. Theslopesandintercepts of theregressions  (dash-dotted lines) are0.682and0.947gmolm'2s'• forFigure 5aand0.308and1.581gmolm'2s'l forFigure 5b.  Although the accuracyof this measurementwas probably not high, it clearly showed the suddenupward dischargeof CO2 stored in the air column correspondingto the unusually large  dropin CO2storage (-13•mol m'2 s'•) (compare Figure3). About 20 sucheventswere observedduring the 1994 and 1996 growingseasons.This phenomenon of lossof CO2to the atmospherein the early morningwas commonlyobservedin a tropical  than0.12ms-•)thanunder stable conditions (mainly nighttime) whenmixingwasweak(Figure5 andTable2). In the lattercase, valuesoftendifferedgreatly,to the degreethat at timestheywere of oppositesign. Generally,therewas slightlybetteragreement between NHE values at the two locations than there was between  the corresponding valuesof Fc. On calm nights,negativeCO2 exchangefrequentlyoccurredat both locationssimultaneously,  rainforestby Graceet al. [1995]. In thatcase,it wasalmostthe suggestingthat thesenegativefluxes were not the resultof measonly time during the day that CO2 was lost to the atmosphere urementerror. Negative fluxes over a relatively large area and abovethe eddy covariancesensorsbecauseduringthe nighttime the persistenceof significantdifferencesbetweenCO2 concentrastabilityabovethe densecanopyappearedto trap respiredCO2. tionsat the two towersfor 6 or more hourssuggestedthe occurHollingeret al. [1998] alsoreportedthatthe CO2storagein the rence of CO2 advectionin the trunk space. Advection flux air column accountedfor the low nighttimeeddy fluxes over a expressedon a groundsurfacearea basiswas calculatedto be as Siberianlarch forest. In contrastto Grace et al. [1995] and Hollargeas2 •molm'2s-1,corresponding to a sustained horizontal linger et al. [1998], on many calm nights in this study, Fc+•a/• did not appearto entirelyaccountfor the likely rateof respiredCO2. This "missing"CO2 is probablyaccountedfor by  massflow [Lee,1998]or CO2storagein theair-filledporesof the soil [Chenet al., 1999].  3.3 Horizontal Variability of the Fluxeswithin the Trunk Space  concentration difference of 3 •molmol'l between thetwotowers anda windspeed of 0.5ms-• atthe4-mheight.These highadvectivefluxeswere consistentwith the abovespatialvariabilityin NHE. This and the considerable short-termhorizontalvariability  in the CO2 concentration at the 4-m height(Figure6) indicated thatthe forestfloor was not asuniformas it appearsto be as con-  cludedby Fan et al. [1995]. The slightlybettercorrelationbe-  tween NHEatthetwolocations (higher r2)forflowparallel tothe  Figure 5 comparesnet hazelnutunderstoryCO2 exchange line connectingthe two towersthan that for perpendicularflow  (NHE, whichis thesumof Fc at the4-m heightandZ•u.• a/at for the 0-4 m air column)obtained duringthe 10-dayperiodat the  also indicatedthe patchinessof the forest floor and understory  with regardto CO2 exchange(Table 2). Nonuniformitywas  two towerlocations.Half-hourlyratesof changein CO2 storage during the daytime were usually lessthan 10% of Fc and about  probablya result of patchinessof the hazelnut,variability in the depthof the surfaceorganiclayer,stonecontentin the glacialtill 20-30% at night. Over thisperiodthe meandaytimeand night- derived mineral soil, and topographicvariability. These results timevaluesof u, at the39.5-mheightwere0.40 and0.26m s'l, showedthat half-hourCO2 exchangeratesmeasuredat one locarespectively. The corresponding valuesat the4-m heightwere tion were usuallynot representativeof a large area of forestfloor 0.06and0.04ms'l [Blanken etal.,1998].Generally, agreement or understory.Over the 1O-dayperiod,however,the averageval-  between NHE valuesat thesetwolocations waspoor;however, there was better agreementin neutral and unstableconditions  ues of NHE  at the two locations  were almost identical:  2.3 and  2.2!.tmol m'2s'l onthe6-mandmaintowers, respectively (Table (mainlydaytime) or whenmixingwasmorevigorous (standard 3). Furthermore,the daytimemeansat the two locationswere deviation of verticalvelocityowat the 4-m heightwasgreater very similar aswas the casefor the nighttimemeans.  YANG ET AL.: VARIABILITY OF COzFLUX IN AN ASPENFOREST  27,659  Table2. Coefficients ofDetermination (rz)UsingDifferent Stratification Criteria intheCorrelation between Fluxes Measuredat Two Locations40 m Apart in the Trunk SpaceEddy Flux Experiment,August12-22, 1994  Time of Day Daytime  t7wat the 4-m Height  u, at the 4-m Height  Nighttime Parallel Perpendicular crw>0.12 ms-1 crw<0.12 ms'1 u,>0.06ms4 n = 196  n = 236  DaytimeWind Direction at 4 m Relative to a Line Connecting theTowers  n = 32  n = 25  n = 98  n = 385  u,<0.06 ms4  n = 155  n = 345  NHE,  gmolm-zS-1  0.113  0.059  0.152a  0.115b  0.290  0.066  0.100  0.069  H, W m'z AE,W m'z  0.424 0.594  0.007 0.018  0.357 0.730  0.022 0.700  0.455 0.328  0.041 0.496  0.507 0.659  0.119 0.521  SeealsoFigures5 and7. aFornighttime, n = 33.  bFornighttime, n = 17. Figure7 compareshalf-hoursensible(H) and latent(/IE) heat riodonthe6-mandmaintowers were18and21 W m'z,respecfluxesin the trunkspaceat the two towers. Unlike COz, storage tively(Table3), whichindicates closeagreement. Analysisalso changesaccountedfor a very small proportionof thesefluxes. showedthatcorrelation between/IEat thetwolocations wasvery Similar to NHE, the sensible heat fluxes measured at the two losimilarfor bothparallelandperpendicular wind directions(Table cations agreed better under neutral and unstable conditions or 2), indicating thattheforestfloor andunderstory weremorehowhenu, atthe4-mheight exceeded 0.06m s-• thantheydidun- mogeneous with regardto/IE thanNHE. For sensibleheatflux, der stableconditions(Figures7a and7b andTable 2). Over the rz washigher forparallel flow(Table2);however, thisdifference becausesensibleheatfluxesabovethe understory 1O-dayperiodthe averagevaluesof H on the 6-m and maintow- is questionable erswere1.6and0.9W m'z,respectively (Table3),whichisclose weregenerallylow [Blankenet al., 1998]. Thesecomparisons suggestthat significantshort-term(e.g., agreementconsideringthat the accuracyof the eddy covariance measurement of H and/IE wasnot betterthan -I-5 W m'z. Half- half hourly)horizontalvariabilityexistsfor COzflux in the trunk hour latent heat fluxes at the two towers were much better correspace all day although it is slightly less during the daytime. lated than NHE  and H in unstable and neutral conditions  were  Short-termhorizontalvariability in sensibleheat flux is lessthan  (Figure7c andTable2). In contrastto NHE andH, the diurnal that for NHE but still very significant. Latentheatflux showsthe patterns of/IE wereverysimilarat thetwo locations on 7 of the least short-termhorizontalvariability especiallywhen mixing is sufficient,which indicatesthe feasibilityof partitioningevapora10 days[Yang,1998]. Averagevaluesof AE overthe 1O-day  520• •  1'1 line  480 A  440  •  •  o  A  A  O  A  360  O  O  32O  320  360  400  440  480  520  6-m scaffoldtower(gmolmo1-1)  Figure6. Comparison ofthehalf-hourly COzconcentrations (ingmolmol'l moist air)measured atthe4-mheight on the6-m scaffoldtowerwith thosecalculatedfor the4-m heighton themaintowerusinga rectangular hyperbolic fit to theconcentrations measured at eightheightsduringAugust12-22, 1994. The trianglesarenighttimedata,and  thecircles aredaytime data.Thedash-dotted lineisthedaytime regression line(C6-m = 0.971Cmain + 15gmolmol4,  rz= 0.88,Syx = 8.8}.tmol mol'l), andthedotted lineisthenighttime regression line(C6. m= 0.695Cmaia + 123}.tmol mol '1,rz= 0.45,S. vx= 19.9gmol mol-1).  27,660  YANG ET AL.' VARIABILITY OF CO2FLUX IN AN ASPEN FOREST Table 3. Comparison of theMeansof ScalarFluxesObtainedon the 6-m ScaffoldTower and the Main Tower in the Trunk SpaceEddy Flux Experiment,August12-22, 1994 DaytimeFlux 6-m Scaffold  NHE,gmolm'2s'1 H, W m'2 AE,W m-2  NighttimeFlux  Main  6-m Scaffold  24-HourFlux  Main  6-m Scaffold  Main  2.0  1.9  2.8  2.7  2.3  2.2  3.1 30.0  2.6 34.0  -0.9 0.5  -1.9 0.5  1.6 18.1  0.9 20.8  tion betweenoverstoryandunderstoryon a 1-2 hour basisduring the daytime. For the 1O-dayperiod,averagingover 5 consecutive dayswas requiredto obtain agreementin the CO2 fluxes at the two locationswithin 10%. This suggests the feasibilityof partitioningCO2uptakebetweenoverstoryand understoryusingaveragesover 5-day periods. In an experimentdesignedto examine the spatialvariabilityof turbulentfluxes abovea uniform even-  morningand late afternoonwere smallcomparedto eddyfluxes F½at the 39.5-mheight. Storagechanges weremainlydueto the lossor gainof CO2belowthe 9-m heightnearsunriseandsunset, whenit accounted for up to 60% of net ecosystem exchange, and to the gainof CO2betweenthe 9- and39.5-mheightsduringthe restof the night. On somecalm nights,CO2 accumulated below the9-m height,whichusuallyresultedin a suddenstrongupward  aged14 m tall loblollypinestand(DukeForest,NorthCarolina) eddyflux of CO2 shortlyafter sunrise. 3. CO2 profile analysisshowed that a minimum of three involving7 towers,Katul et al. [1999] foundthat F½was also spatiallyheterogeneous.In contrastto our study,they found that heightswasrequiredto estimatetheratein changeof CO2storage ZE wasas spatiallyheterogeneous asF½,whileH wasrelatively in the 0- to 39.5-m air column within 5% of values obtained ushomogeneous. ing concentrations measuredat eightheights. The threeheights had to includeone heightaboveand two heightsbelowthe base of the aspencanopy.  4. Conclusions  4.  The correlation between fluxes measured at two locations  1. During the leaflessand full-leaf perind,q in a 21.5-m tall beneaththe aspencanopyabouttwo tree heightsapartincreased aspenforest,verticalCOz concentration gradientsin the 9- to in thefollowingorder:F½,H, and/IE. The low correlationin the 39.5-m layerwere smallduringthe daytimeand nighttime. Be- caseof Fc, especiallyin stableconditions,indicatessignificant (e.g.,halfhour)horizontalvariability.However,when low the 9-m heightthere were significantverticalgradientsand short-term markeddiurnalvariationsparticularlyaroundsunriseandsunset. averagedover a period of 5 days, daytimeand nighttimeCO2 2. Changesin COz storagein the air columnbeneaththe fluxesat the two locationsagreedto within 10%, suggesting that above-forest eddy covariancesystem(39.5 m) betweenlate partitioningnet ecosystemexchangebetweenthe overstoryand ,  60  20  40  10  20 -10 -20  E -20  g -?40  0  2O  4O  6O  o  -3-030 -20 -10 0 40  ,- 150  10 20  !  (d) stable v  E 100  2E  20  50 v  vv  v Vr2= 0.09  -5-050 0  50 100 150  2)  0  20  40  6-m scaffold tower (W m-2) Figure 7. Sameas Figure5 exceptfor sensible(H) and latent(AE) heatfluxes. The solidlinesare one-to-one  lines. Theslopes andintercepts oftheregressions (dash-dotted lines) are:(a)0.741,0.724W m'z,(b)0.280,-1.495  W m'z,(c)0.684,16.943W m'z,and(d)0.363,0.867W m'2.  YANG ET AL.: VARIABILITY OF CO2FLUX IN AN ASPEN FOREST understory/soil can be doneusingaveragesover periodsat least5 dayslong,but not on a half-hourlybasis. Acknowledgment. The authorsfrom UBC gratefullyacknowledge fundingprovidedby a 4-year Natural Scienceand EngineeringResearch  Council(NSERC)Collaborative SpecialProjectGrantin supportof BOREAS, an NSERC OperatingGrant (TAB), and grants from the  27,661  Greco, S., and D. D. Baldocchi, Seasonalvariations of COz and water va-  por exchangeratesovera temperatedeciduous forest,Global Change Biol., 2, 183-197, 1996.  Hollinger,D. Y., F. N. Kelliher,J. N. Byers,J. E. Hunt,T. M. McSeveny, andP. L. Weir, Carbondioxideexchangebetweenan undisturbedoldgrowthtemperateforestand the atmosphere, Ecology, 75, 134-150, 1994.  Hollinger,D. Y., et al., Forest-atmosphere carbondioxide exchangein AES/NSERC Joint ScienceSubventionProgramandthe CanadianForest easternSiberia,Agric. For. Meteorol.,90, 291-306, 1998. Service. We greatlyappreciatethe supportgivenby numerousindividuJarvis,P. G., J. M. Massheder,S. E. Hale, J. B. Moncrieff,M. Rayment, als. On-siteassistance was providedby JohnDeary, Tom Hertzog, and and S. L. Scott, Seasonalvariationof carbondioxide, water vapour Monica Eberle. Field operationswere made possibleby the supportof and energyexchanges of a borealblack spruceforest,J. Geophys. PrinceAlbert National Park personnel,especiallyMary Dahlman, Paula Res., 102, 28,953-28,966, 1997. Pacholek,and Murray Heap. JanuszOlejnik and Marian Breazuassisted Katul, G., et al., Spatialvariabilityof turbulentfluxes in the roughness with instrumentconstructionand electronics. Gerry den Hartog, Zoran sublayerof an even-agedpine forest,BoundaryLayer Meteorol., in Nesic,Xuhui Lee, Ralf Staebler,Alan Barr, Joe Eley, Rick Ketler, Uwe press,1999. Gramann,Craig Russell,AishengWu, hobel Simpson,Grant Edwards, Lee, X., On micrometeorological observations of surface-airexchange andJoseFuentesprovidedvariousformsof field assistance.SiguoChen overtall vegetation, Agric. For. Meteorol.,91, 39-49, 1998. and Wenjun Chen assistedus both in the field and with data analysis. Lee, X., T. A. Black, and M.D. Novak, Comparisonof flux measureWe greatly appreciatethe helpful commentsand suggestions made by mentswith open-and closed-path gasanalyzersabovean agricultural threeanonymous reviewers. field and a forest floor, BoundaryLayer Meteorol., 67, 195-202, 1994.  References Baldocchi,D. D., Flux footprintswithin and over forestcanopy,Boundary LayerMeteorol.,85, 273-292,1997. Baldocchi, D. D., and B. A. Hutchison,Turbulencein an almond orchard:  Vertical variationsin turbulentstatistics,BoundaryLayer Meteorol.,  Lee, X., T. A. Black, G. den Hartog, H. H. Neumann,Z. Nesic, and J. Olejnik, Carbondioxide exchangeand nocturnalprocesses over a mixeddeciduous forest,Agric. For. Meteorol.,81, 13-29, 1996. Lee, X., J. D. Fuentes,R. M. Staebler,and H. H. Neumann,Long-term observation of the atmospheric exchangeof COz with a temperatedeciduousforestin southernOntario,Canada,J. Geophys.Res.,in press, 1999  40, 127-146, 1987.  Schuepp,P. H., M. Y. Leclerc,J. I. MacPerson,and R. L. Desjardins, Baldocchi,D. D., andT. P. Meyers,Tracegasexchangeabovethe forest Footprintpredictionof scalarfluxesfrom analyticalsolutionsof the floorof a deciduous forest,1, Evaporation andCOz flux, J. Geophys. diffusionequation, BoundaryLayerMeteorol.,50, 355-373,1990.  Res., 96, 7271-7285, 1991. Tanner, C. B., and G. W. Thurtell, Anemoclinometermeasurementsof Baldocchi,D. D., and C. Vogel, A comparativestudyof watervapor,enReynolds stressandheattransport in theatmospheric boundarylayer, ergy and COz flux densitiesaboveand below a temperatebroadleaf Res.Dev. Tech.œep.ECOM-66-G22F,Univ. of Wis., Madison,1969. anda borealpineforest,TreePhysiol.,16, 5-16, 1996. Verma, S. B., D. D. Baldocchi, D. E. Anderson,D. R. Matt, and R. E. Baldocchi,D. D., C. Vogel, andB. Hall, Seasonalvariationof carbondiClement,Eddy fluxesof COz, watervaporand sensibleheat over a oxideexchangeratesaboveandbelowa borealjack pineforest,Agric. deciduous forest,BoundaryLayerMeteorol.,$6, 71-97, 1986. For. Meteorol., 83, 147-170, 1997. Webb, E. K., G.I. Pearman,and R. Leuning,Correctionof flux measBlack, T. A., et al., Annual cycle of water vapor and carbon dioxide urements for densityeffectsdueto heatandwatervaportransfer,Q. J. abovea borealaspenforest,Global ChangeBiol., 2,219-229, 1996. œ. Meteorol. Soc., 106, 85-100, 1980. Blanken, P. D., Evaporationwithin and above a boreal aspenforest, Wofsy,S.C., M. L. Goulden,J. W. Munger,S. -M. Fan,P.S. Bakwin,B.  Ph.D. thesis, Univ. of B.C., Vancouver, Canada, 1997.  Blanken,P. D., T. A. Black, H. H. Neumann,G. den Hartog,P. C. Yang, Z. Nesic, R. Staebler, W. Chen, and M.D.  Novak, Turbulent flux  C. Daube,S. L. Bassow,andF. A. Bazzaz,Net exchangeof COz in a mid-latitude forest,Science,260, 1314-1317, 1993.  Yang,P. C., Carbondioxideflux withinandabovea borealaspenforest, measurements aboveandbelowthe overstoryof a borealaspenforest, Ph.D. thesis,Univ. of B.C., Vancouver,Canada, 1998. BoundaryLayerMeteorol., 89, 109-140,1998. Chen,W. J., T. A. Black, P. C. Yang, A. G. Barr, H. H. Neumann,Z. NeT. A. Black, M.D. Novak, and P. C. Yang, Facultyof Agricultural sic, M.D. Novak, J. Eley, R. J. Ketler, and R. Cuenca,Effectsof cliUniversityof BritishColumbia,139-2357Main Mall, Vancoumaticvariabilityon the annualcarbonsequestration by a borealaspen Sciences, ver, British Columbia, Canada V6T 1ZA. (ablack@interchange.ubc.ca; forest,GlobalChangeBiol., 5, 41-53, 1999. chenggan @interchange.ubc.ca.) Fan, S. -M., M. L. Goulden,J. W. Munger,B.C. Daube,P.S. Bakwin, S. novk@interchange.ubc.ca; P. D. Blanken, University of Colorado, Boulder, CO 80302. C. Wofsy, J. S. Amthor,D. R. Fitzjarrald,K. E. Moore, and T. R. @spot.colorado.edu.) Moore, Environmentalcontrolson the photosynthesis and respiration (blanken H.H. Neumann,Atmospheric Environment Service,Downsview,Onof a boreallichen woodland:A growingseasonof whole-ecosystem exchangemeasurements by eddy correlation,Oecologia,102, 443- tario,CanadaM3H 5T4. (hneumann@ec.gc.ca.) 452, 1995.  Grace,J., et al., Carbondioxideuptakeby an undisturbedtropicalrain forest in southwest Amazonia, 1992 to 1993, Science, 270, 778-780, 1995.  (ReceivedSeptember 15, 1998;revisedApril 21, 1999; accepted April 28, 1999.)  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items