Quantifying the effect of slope on extensive green roof stor

非常好的资料

ecologicalengineering31(2007)225–231

Quantifyingtheeffectofslopeonextensivegreenroofstormwaterretention

KristinL.Gettera, ,D.BradleyRowea,JeffreyA.Andresenb

ab

DepartmentofHorticulture,MichiganStateUniversity,EastLansing,MI48824,UnitedStatesDepartmentofGeography,MichiganStateUniversity,EastLansing,MI48824,UnitedStates

article

Articlehistory:

infoabstract

Impervioussurfaces,suchasrooftops,parkinglots,androads,increaserunoffandthepotentialfor ooding.Greenrooftechnologies,whichentailgrowingplantsonrooftops,areincreasinglybeingusedtoalleviatestormwaterrunoffproblems.Toquantifytheeffectthatroofslopehasongreenroofstormwaterretention,runoffwasanalyzedfrom12exten-sivegreenroofplatformsconstructedatfourslopes(2%,7%,15%,and25%).Raineventswerecategorizedaslight(<2.0mm)(0.08in.),medium(2.0–10.0mm)(0.08–0.39in.),orheavy(>10.0mm)(>0.39in.).Datademonstratedanaverageretentionvalueof80.8%.Meanreten-

Received12February2007Receivedinrevisedform31May2007

Accepted19June2007

Keywords:VegetatedroofRunoffEco-roof

tionwasleastatthe25%slope(76.4%)andgreatestatthe2%slope(85.6%).Inaddition,runoffthatdidoccurwasdelayedanddistributedoveralongperiodoftimeforallslopes.Curvenumbers,acommonmethodusedbyengineerstoestimatestormwaterrunoffforanarea,rangedfrom84to90,andarealllowerthanaconventionalroofcurvenumberof98,indicatingthatthesegreenedslopesreducedrunoffcomparedtotraditionalroofs.

©2007ElsevierB.V.Allrightsreserved.

1.Introduction

Impervioussurfacescontinuetoexpandasweconstructbuildings,roads,andparkinglots.IntheUnitedStates,itisestimatedthat10%ofresidentialdevelopmentsand71–95%ofindustrialareasandshoppingcentersarecoveredwithimper-vioussurfaces(Ferguson,1998).Two-thirdsofallimperviousareaisintheformofparkinglots,driveways,roads,andhigh-ways(WaterResourcesGroup,1998).

Coveringnaturalsurfacescausesmanyproblems.Greaterrunoff(Scholz-Barth,2001)increasesthepotentialfor ooding,reducesin ltrationintothegroundwatersystem(Barnesetal.,2001),andcanpotentiallycontaminatesurfacewatersduetoparticulatematterintherunoff(USEPA,1994;Ferguson,1998).Otherproblemswithimpervioussurfacesincludehigherambientairtemperatures(USEPA,2003),increasednoise,poorerairquality(LieseckeandBorgwardt,1997;YokTanandSia,2005),andalossofbiopersity(Bastinetal.,1999).

Greenroofsareonepotentialremedyfortheseproblems.Establishingplantmaterialonrooftopsprovidesnumerousecologicalandeconomicbene ts,includingstormwaterman-agement,energyconservation,mitigationoftheurbanheatislandeffect,increasedlongevityofroo ngmembranes,andmitigationofnoiseandairpollution,aswellasamoreaesthet-icallypleasingenvironmentinwhichtoworkandlive(GetterandRowe,2006;Liesecke,1998;LiuandMinor,2005GRHC;MengandHu,2005;SimmonsandGardiner,2007;VanWoertetal.,2005;VillarrealandBengtsson,2005).

Manyconsiderthereductionofstormwaterrunofftobethegreatestenvironmentalservicethatgreenroofsprovide.Inagreenroofsystem,muchoftheprecipitationiscapturedinthemediaorvegetationandeventuallyevaporatesfromthesoilsurfaceorisreleasedbackintotheatmospherebytranspiration.Whilethechosentypeofgreenroofsystem(design,substratedepth,andplantspecies)willaffectreten-tion,researchhasshownreductionsof60–100%inrunoff

Correspondingauthor.Tel.:+15173555191x1341.E-mailaddress:smithkri@msu.edu(K.L.Getter).

0925-8574/$–seefrontmatter©2007ElsevierB.V.Allrightsreserved.doi:10.1016/j.ecoleng.2007.06.004

非常好的资料

226

ecologicalengineering31(2007)225–231

(Liesecke,1998;Moranetal.,2004;DeNardoetal.,2005;VanWoertetal.,2005).

Sincegreenroofsretainstormwater,theycanmitigatetheeffectsofimpervioussurfacerunoff.Peck(2005)estimatedthatif6%ofallbuildingsinTorontohadgreenroofs,itwouldresultinthesamestormwaterretentionimpactasbuildinga$60million(CDN)storagetunnel.Likewise,inWashington,DC,if20%ofallbuildingsthatcouldsupportagreenroofhadone,theywouldaddover71millionliters(19milliongallons)tothecity’sstormwaterstoragecapacityandstoreapproxi-mately958millionliters(253milliongallons)ofrainwaterinanaverageyear(Deutschetal.,2005).

InGermany,tworesearchersfoundnosigni cantdiffer-enceinretentionamountsacrossdifferentlyslopedroofs(Liesecke,1999;Schade,2000),whileotherscientistsareestablishingdifferences(VanWoertetal.,2005;VillarrealandBengtsson,2005).Thecontradictingresultsmaybeduetorainfallpatternsatdifferentlocales.Rainfallintensity,dura-tion,andinitialsubstratemoisturecontentallin uenceretention.Drysubstrateconditionspriortorainfallresultingreaterstormwaterretentioncomparedtoinitiallywetcondi-tions(VillarrealandBengtsson,2005;ConnellyandLiu,2005).Environmentaldifferencesmayalsoin uencethechoiceforsubstratedepthandplantmaterial,whichmayinturnin u-encestormwaterretention.

Acommonandwidespreadmethodforestimatingstormwaterrunoffforaregionorareaisthecurvenumber(CN)methoddevelopedbytheUSDASoilConservationServices(USDASCS),nowtheUSDANaturalResourcesConservationService(USDANRCS).ThismethodstatestherelationshipbetweenrainfallandrunoffwiththeequationF/S=Q/P,whereFistheactualretention(P Q),Sthepotentialretention,Qtheactualrunoff,andPthepotentialrunoffortotalrainfall(NRCS,2004).Thepotentialretention(S)canthenbeconvertedtoacurvenumberwiththeformulaCN=25,400/(254+S)whereSisinmm(Hawkins,1993).Curvenumbersaredimensionlessandrangefrom0(norunoff)to100(allprecipitationresultsinrunoff).AllimpervioussurfacessuchaspavedroadsandconventionalroofsareassignedaCNof98(NRCS,2004).

Sincegreenroofsaremorefrequentlybeingusedasatoolformanagingstormrunoff,theobjectiveofthisstudywastoquantifytheeffectofslopeonstormwaterretentionanddevelopcurvenumbersforgreenroofsatfourdifferentslopes.

2.

Materialsandmethods

2.1.

Greenrooftestingplatforms

Twelveroofplatformswithdimensionsof2.44m×2.44m(8.0ft×8.0ft)wereconstructedbyChristenDETROITRoo ngContractors(Detroit,MI)attheMichiganStateUniversityHor-ticultureTeachingandResearchCenter(EastLansing,MI).Eachplatformreplicatedacommercialextensivegreenroof,includinginsulation,protective,andwaterproo ngmem-branelayers.ConstructiondetailsareoutlinedinVanWoertetal.(2005).

Aluminumsheetmetaltroughswereattachedonthelowendoftheplatformstodirectstormwaterrunoffthroughthemeasuringdevicesusedtoquantifyrunoff.Thewood-framed

platformsincludedsidesthatextended20.3cm(8in.)abovetheplatformdeck,alsocoveredwithwaterproo ngmem-brane.Allplatformswereplacedwiththelowendoftheslopefacingsouthtomaximizesunexposure.

2.2.Drainagesystemandvegetationcarrier

EachplatformwascoveredwithaXeroFlorXF108drainagelayer(WolfgangBehrensSystementwicklung,GmbH,GroßIppener,Germany)installedoverthewaterproo ngsystem,whichallowedexcesswaterto owofftheroof.Foradditionalwaterholdingcapacity,a0.75cm(0.26in.)thickmoistureretentionfabric(XeroFlorXF159)capableofretainingupto5.92kgm 2ofwaterwasplacedoverthedrainagelayer.Abovetheretentionfabricwasthevegetationcarrier(XeroFlorXF301).

2.3.Plantestablishment

Growingsubstrate(Table1)wasplacedontopofthevegeta-tioncarrieratadepthof6.0cm(2.4in.).Thewaterretentionfabricandsubstratetogetherhavethepotentialtoholdupto12.0mm(0.5in.)ofrainfall.SeedsweresownandestablishedonthegrowingsubstrateperVanWoertetal.(2005).SpeciesincludedSaxifragagranulataL.(meadowsaxifrage),SedumacreL.(bitingstonecrop),SedumalbumL.(whitestonecrop),SedumkamtschaticumellacombianumFisch.(kamtschatkastonecrop),

Table1–InitialphysicalandchemicalpropertiesofsubstrateComponent

Unit

Totalsand(%)

91.18Verycoarsesand(1–2mm)(%)21.96Coarsesand(0.5–1mm)(%)

40.80Mediumsand(0.25–0.5mm)(%)24.66Finesand(0.10–0.25mm)(%)

3.36Very nesand(0.05–0.10mm)(%)0.40Silt(%)5.60Clay(%)

3.22Bulkdensity(g/cm3)1.16Porespace(%)

41.41Air lledporosity(%)

21.43Waterholdingcapacityat0.01MPa(%)17.07pH

7.9Conductivity(EC)(mmho/cm)3.29Nitrate(ppm)

203Phosphorus(ppm)65.8Potassium(ppm)622Calcium(ppm)214Magnesium(ppm)60Sodium(ppm)164Sulfur(ppm)184Boron(ppm)0.5Iron(ppm)

9.0Manganese(ppm)15.7Zinc(ppm)5.7Copper(ppm)

0.6

AnalysisperA&LGreatLakesLaboratories,Inc.,Ft.Wayne,Indiana.

非常好的资料

ecologicalengineering31(2007)225–231

227

SedumpulchellumMichx.(bird’sclawsedum),Sedumre exumL.(crookedstonecrop),SedumspuriumBieb.‘Coccineum’(creep-ingsedum),andSedumspuriumBieb.‘SummerGlory’(creepingsedum).Fullcoveragewasachieved(nosubstrateexposed)andmaintainedsinceJuly2002.

2.4.Treatments

Platformsweresetatoneoffourslopes(2%,7%,15%,and25%)inacompletelyrandomizeddesign(CRD)witheachslopereplicatedthreetimes.Platformswereadjustedtotheappro-priateslopeinApril2005.Becauseaslopedplatformreducesthehorizontalareauponwhichrainfalls,theeffectiveareaofeachplatformwascalculatedbasedonslopeandoriginalplatformarea.Thus,theeffectiveplatformareaswere5.49m2(59.07ft2),5.48m2(58.94ft2),5.43m2(58.41ft2),and5.32m2(57.26ft2)forthe2%,7%,15%,and25%treatments,respec-tively.

Becausethisstudyutilizedroofplatformsthatwerethreeyearsoldatthebeginningofthestudy,substratesamplesweretakenattheconclusionofthisstudyinordertoquan-tifysubstratechangesovertime.Soilcores(13.0cm(5.1in.))weretakenatthreerandomplacesamongstthetwelveplat-formsandwereanalyzedfororganicmatter(lossonignitionat550 C),porespace,freeairspace,andwaterholdingcapacity(A&LGreatLakesLaboratories,FortWayne,IN).Theseresultswerecomparedwithpreviousanalysisoffreshsubstrate.

2.5.Datacollection

RainfallandrunoffwererecordedonaCR10Xdatalogger(CampbellScienti c,Inc.,Logan,UT)thatwasplacedalong-sidethreeAM16Tmultiplexers.TwelveTE525WStippingbucketraingages(CampbellScienti c,Inc.,Logan,UT)wereeachsituatedunderneathaplatformtocollectrunofffromthealuminumtroughsanda13thtippingbucketmeasuredrain-fall.Thetwelverunofftippingbucketswerecoveredwithaplastic10inroundplantsaucerthataccommodatedaholetoallowwaterfromthealuminumtroughtoentertheraingagewhilealsoexcludingrainfall.Accuracyoftheraingageswasreportedbythemanufacturertobe±1%,+0and 2.5%,and+0and 3.5%forrainfallsof<25.4mmh 1,25.4–50.8mmh 1,and50.8–76.2mmh 1,respectively.

Datawererecordedcontinuouslyfrom26April2005until1September2006.Thedataloggerwasprogrammedtocollectvalueseveryminuteandtotalswereputoutevery5min,24hadaythroughouttheperiod.Dataweredownloadedoffthedataloggerandontoalaptopcomputereveryweek.

2.6.Dataanalysis

Retentiondatawereanalyzedfromallraineventsthatoccurredduringtemperaturesabove0 C(32 F)asapercent-ageoftotalrainfallforeachrainevent.Inordertoexcludemeltingprecipitationinrunoffdata,analysiswaslimitedtodatesbetween26April2005and22November2005andbetween12April2006and1September2006.Retentionisde nedhereasprecipitationthatdidnotrunofftheplat-forms.Independentraineventswerede nedasprecipitationeventsseparatedby6ormorehours.Intheeventrunoffwas

stilloccurring6hafterthe rstevent,thetwoeventswerecombined.Raineventswerearbitrarilycategorizedbyrelativeintensityaslight(<2.0mm)(0.08in.),medium(2.0–10.0mm)(0.08–0.39in.),orheavy(>10.0mm)(>0.39in.).Therangeofeachcategorywaschosentoobtainraineventsamplesizesthatweresimilaracrossallthreecategories.Therewasatotalof62rainevents.

Datawereanalyzedtwoways.Inthe rst,meanper-centretentionperraineventwasanalyzedusinganANOVAmodelwithroofslopeandrainfallcategoryas xedeffects.Althoughoriginalmeansarepresented,allretentionvaluesweretransformedpriortoanalysisusinganarcsinesquareroottransformationtostabilizethevarianceandnormal-izethedataset(Underwood,1998).Signi cantdifferencesbetweentreatmentsweredeterminedusingmultiplecompar-isonsbyLSD(PROCMIXED,SASversion8.02,SASInstitute,Cary,NC).Thesecondanalysiswastodeterminecurvenum-bersforeachgreenroofslopebyregressingforSintheformulaF/S=Q/P(PROCREG,SASversion8.02,SASInstitute,Cary,NC)andthenconvertingStoacurvenumberwiththeequationCN=25,400/(254+S)whereSisinmm(CarterandRasmussen,2006).

3.Resultsanddiscussion

Duringthestudytherewere94dayswithquanti ableprecip-itation,resultinginatotalof62raineventsthatwereusedinanalysis(Fig.1).Themaximumprecipitationfor1daywas38.1mmduringthestudy,whileamaximumsingleraineventexceeded40mm(Fig.2).Rainfallwasdistributedas16light(<2mm),24medium(2–10mm),and22heavy(>10mm)rainevents(Fig.2).Dailyminimumambientairtemperaturesdur-ingthedatacollectionperiodrangedfrom 6.7 Cto25.3 C(19.9–77.5 F)anddailymaximumambientairtemperaturesrangedfrom 2.3 Cto34.8 C(27.9–94.6 F)(Fig.1).

TheANOVAmodelshowedraincategoryandslope,aswellastheinteractionofboth,tobesigni cant(Table2).Represen-tativehydrographs(Fig.3)andcumulativehydrographs(Fig.4)illustratetheeffectofroofslopeonquantityofrunoffandoveralldelayforlight,medium,andheavyrainevents.Initialrunoffdelayfortheseraineventsisminimalforallslopes.ThiscontradictsDeNardoetal.(2005),VanWoertetal.(2005),andCarterandRasmussen(2006)whoreported4h,40min,and34mininitialdelays,respectively.Perhapstheintensityoftherainfallormoistureconditionofthesubstratepriortothesestormeventsexplainsthedifference.Anotherexplanationforthesecontradictingresultsisthatwiththeexceptionofthe2%slope,allslopesevaluatedinthisstudyaresteeperthanDeNardoetal.(2005),VanWoertetal.(2005),andCarterandRasmussen(2006)andthisstudyalsohasashallowermediadepththanDeNardoetal.(2005)andCarterandRasmussen(2006).

Inaddition,thisroofecosystemhadbeenestablishedfor3yearspriortocommencingthisstudy,whichisolderthanallofthepreviouslymentionedstudies.Thisgreatermatu-ritymayeffectthehydraulicconductivityofthesubstrate.Mentensetal.(2006)indicatedthatroofagewasnotcorrelatedtothequantityofretention,butroofagemayaffectthetimepatternofretention.Overallporespaceandchangesinpore

非常好的资料

228

ecologicalengineering31(2007)

225–231

Fig.2–Frequencyofraineventsincludedinthestudyfrom26April2005to22November2005and12April2006to1September2006.Rainfallmeasurementsweretakenfromtippingbucketraingaugesmountedattheresearchsite.

Fig.1–Dailymaximumandminimumtemperatures( C)andprecipitation(mm)throughoutthestudy(1April2005to1September2006).DataarefromtheMichiganAutomatedWeatherNetwork’sEastLansingweatherstationlocatedadjacenttotheresearchsite.

sizemayoccurovertimeasaresultofsettlingorasaresultofchangesinorganicmattercontent.Inthisstudy,maturesubstrateexhibitedgreatervaluesforporosity,freeairspace(macropores),organicmatter,andwaterholdingcapacityattheconclusionofthisstudyrelativetotheinitialsubstrate(Table3).Increasedfreeairspace,resultingfromchannelsformedbydecayingrootsorburrowinginsects,mayincreasepreferentialmacropore owthroughthesubstrate,therebyresultinginquickerinitial

runoff.

Runoffwasspreadoutovertimeacrossalltreatmentswiththe nalrunofflasting4h20min,10h45min,and13h45minforlight,medium,andheavyrainevents,respectively,afterrainfallstopped.TheseresultsaresimilartoVanWoertetal.(2005),Liu(2003),andMoranetal.(2004).LiuandMoranalsofoundthatthisdelayedrunoffwasatalower owrate.Byslowingdowntherateofrunoffandreleasingitoutoveralongerperiodoftime,greenroofscanhelpmitigatetheero-sionalpowerofrunoffthatdoesenterstreams,eitherthroughdirectrunofforstormsewers.Itcanalsopreventcombinedstormwatersewersystemsfromover owing,byallowingittoprocessrunoffforalongertimeatalower owrate.Theseresultsmayin uencestormwatermanagementpracticesordesignofmunicipalstormwaterandsewagesystems.

Thegreenroofsretainedanaverageof80.2%ofprecipi-tationaveragedacrossallslopesandraincategories(Table4).Meanretentionwasleastatthe25%slope(75.3%)andgreatestatthe2%slope(85.2%).Inaddition,retentionvaluesdecreasedasslopeincreased.Retentionvalueswerehighestforlightrainevents(94.2%)andlowestforheavyrainevents(63.3%).

Table2–ANOVAtableforrainfallretentionoverthe2-yearperiod(26April2005to22November2005and12April2006to1September2006)fromfourroofplatformtreatmentsreplicatedthreetimesSourceofvariation

Model

RaincategoryaSlopeb

Category×slopeError

Correctedtotal

Degreesoffreedom

11236598609

Sumofsquares

80.574.61.52.663.3143.8

Meansquares

7.337.30.50.40.1

F-Statistic

69.1352.24.74.1

P-value

<.0001<.0001.0029.0005

Retentionisthedependentvariable.Roofslopeandraincategoryareindependentvariables.

ab

Vegetatedroofplatformssetat2%,7%,15%,and25%slopewith6.0cm(2.4in.)ofsubstrate.

Raineventcategorieswerelight(<2.0mm)(0.08in.)(n=16),medium(2.0–10.0mm)(0.08–0.39in.)(n=24),heavy(>10.0mm)(>0.39in.)(n=22),andoverall(n=62).

非常好的资料

ecologicalengineering31(2007)225–231

229

Fig.3–Runoffhydrographsofselectedrepresentativerainevents:(A)heavy(23.37mm)(0.92in.),(B)medium

(5.08mm)(0.2in.),and(C)light(1.02mm)(0.04in.).Linesrepresentrunoff(mm)froma2%,7%,15%,or25%roofslopewith6.0cm(2.4in.)ofmedia.Valuesareaveragesofthreereplicationsmeasuredusingtippingbucketraingaugesmountedattheresearchsite.

Fig.4–Cumulativehydrographsofselectedrepresentativerainevents:(A)heavy(23.37mm)(0.92in.),(B)medium(5.08mm)(0.2in.),and(C)light(1.02mm)(0.04in.).Linesrepresentrunoff(mm)froma2%,7%,15%,or25%roofslopewith6.0cm(2.4in.)ofmedia.Valuesareaveragesofthreereplicationsmeasuredusingtippingbucketraingaugesmountedattheresearchsite.

Thisdemonstratesthatthesubstratehasalimitedstoragecapacity;onceitissaturatedtheprecipitationrunsoff.

RetentiondataagreewithVanWoertetal.(2005)andCarterandRasmussen(2006).However,ourretentionvaluesaremuchhigherthanDeNardoetal.(2005),Liesecke(1998),andMentensetal.(2006)whoallreportedanaverageof45%,40–50%,and45%retention,respectively.Thismaybedue

to

differencesinsubstratedepth,antecedentsubstratemois-turestatus,slope,orprecipitationpatterns.Butitismostlikelyduetothefactthatallofthelatterresearchersusedlargestormsintheirstormwatertesting.Inaddition,Liesecke(1998)andMentensetal.(2006)followedFLL(http://www.f-l-l.de/english)guidelineswhichemploynearlysaturatedantecedentmoistureconditionsfollowedbyasimulatedrainthatconstitutesa100-yearstorm.Thisisverydifferentfrom

Table3–Organicmattercontentandphysicalpropertiesofinitialsubstratepriortoplanting(2002)andafter5yearsonagreenroof(2006)Sample

InitialsubstrateMaturesubstrate

Organicmatter(%)

2.334.25

Porespace(%)

41.4181.84

Freeairspace(%)

21.4314.40

Waterholdingcapacity(%)

17.0767.44

AnalysisperA&LGreatLakesLaboratories,Inc.,Ft.Wayne,Indiana.

非常好的资料

230

ecologicalengineering31(2007)225–231

Table4–Meanpercentage±thestandarddeviationoftotalrainfallretentionoverthe2-yearperiod(26April2005–22November2005and12April2006–1September2006)fromfourroofplatformtreatmentsreplicatedthreetimesTreatmenta

2%7%15%25%Overall

ab

Lightb(%)

93.394.094.095.5

±±±±3.4bAc3.1cA3.2cA2.9cA

Medium(%)

92.289.588.687.8

±±±±

9.5bA12.7bA13.3bA16.3bA

Heavy(%)

71.466.458.457.1

±±±±

18.1aC18.5aB17.4aA16.1aA

Overall(%)

85.282.278.075.3

±±±±15.9B18.3AB21.0A22.3A

94.2±3.3c89.5±12.8b63.3±18.4a80.2±19.6

c

Retentionfromvegetatedroofplatformssetat2%,7%,15%,and25%slopewith6.0cm(2.4in.)ofsubstrate.

Raineventcategorieswerelight(<2.0mm)(0.08in.)(n=16),medium(2.0–10.0mm)(0.08–0.39in.)(n=24),heavy(>10.0mm)(>0.39in.)(n=22),andoverall(n=62).

Meanseparationinrows

andcolumnsbyLSD(P≤0.05).Lowercaselettersdenotecomparisonsacrossraincategorieswithininpidualslopes(n=12).Uppercaselettersincolumnsdenotedifferencesamongslopes(n=12).

ournaturalconditions,whichwouldhavevaryingantecedentmoistureconditionsandvaryingstormvolumes.Forexam-ple,on18June2006,42.2mm(1.7in.)ofprecipitationoccurredfollowinganearlierraineventof7.62mm(0.3in.).Theserela-tivelydryantecedentconditionsretained68%,64%,57%,58%for2%,7%,15%,and25%slopes,respectively.Incontrast,on28August2006,28.7mm(1.1in.)ofrainfellafterarainjust1.5daysbeforeof8.1mm(0.32in.).Therelativelywetantecedentconditionsretained45%,30%,27%,29%for2%,7%,15%,and25%slopes,respectively.

Inaddition,organicmattercompositionandageofthesub-stratemayaffectretentionvolumesaswell.WhilereportsfromGermany(Mentensetal.,2006)indicatethatroofagedoesnotaffectthequantityofretention,our5-year-oldsubstratehadnearlytwicethewaterholdingcapacityasnewsubstrate(Table3).Increasesinorganicmatterandmicroporesmayincreasewater-holdingcapacity,whichincreasestotalreten-tion,butincreasedmacropores(channels)alsoreducedtheinitialdelay.

Schade(2000)andLiesecke(1998)concludedthatgreenroofslopedidnotaffectretentionamountsforslopesrangingfrom2%to58%.Ourresultsarecontradictoryinthattheeffectof

roofslopewassigni cantwhencomparing2%and15%slopes,aswellas2%and25%slopes.Thisdifferenceagainisproba-blyduetotheseresearchersusingwetantecedentmoistureconditionsand100-yearstormwatervolumesinsimulatedconditions,whichisdifferentfromourstudydesign.Maybefora100-yearstormevent,slopedoesnotin uenceretention,butfornormalraineventsitdoes.Greenroofswillfunctionthemajorityofthetimeundernormalweatherconditions,notin100-yearstormswherethesubstrateisinundatedwithwater.Curvenumberswerecalculatedtobe84,87,89,and90for2%,7%,15%,and25%slopes,respectively.Allofthesenumbersarelowerthanaconventionalroofcurvenumberof98,indi-catingthatallofthesegreenedslopeshadlessrunoffthantraditionalblackroofs(Fig.5).ThisagreeswithVanWoertetal.(2005)whocomparedconventionalgravelballastedroofswithgreenroofsandfoundthattraditionalroofsretainedtheleastrainfall.Curvenumbersalsoincreaseinvalueasslopeincreased,indicatingmorerunoffasslopesbecamesteeper(Fig.5).UsingthesecurvenumbersintheequationCN=25,400/(254+S)andsolvingforpotentialretention(S)we ndthatSrangesfrom28.2mmto48.4mm(1.1–1.9in.).These ndingsaresimilartoCarterandRasmussen(2006)whofoundacurvenumberof86(S=40.5mm)(S=1.6in.)foragreenroofwith<2%slopeand7.62cm(3.0in.)ofsubstrate.Otherlandandsurfacecovertypeswhichhavethesimilarcurvenum-berstotheserangefromclaysoilpasturesinfairconditiontogravelroadsatopclaysoil(NRCS,2004).Thesecurvenumberswillassistengineersandstormwatermanagersinestimatingstormwaterrunoffpeakratesandrunoffquantitiesoflargerwatershedsthatimplementgreenroofs.

4.Conclusion

Fig.5–Curvenumbers(CN)fromvegetatedroofplatformssetat2%,7%,15%,and25%slopeswith6.0cm(2.4in.)ofsubstrateoverthe2-yearperiod(26April2005–22

November2005and12April2006–1September2006)fromfourroofplatformtreatmentsreplicatedthreetimes.

Thisstudydemonstratedthatgreenroofslopedoeshaveaneffectonrunoffretentionquantities.Retentionvaluesdecreasedasslopeincreasedandwassigni cantforslopesbetween2%and15%aswellasbetween2%and25%.Inaddition,greenroofcurvenumberswereshowntobemuchlowerthantraditionalroo ngmaterials,whicharetypicallyassignedacurvenumberof98.Inthisstudy,curvenumbersrangedfrom84to90,resultinginapotentialretention(S)rangingfrom28.2mmto48.4mm(1.1–1.9in.).

TheseconclusionsareapplicabletothemidwesternUnitedStatesandothergeographicalareaswithsimilarclimates.

非常好的资料

ecologicalengineering31(2007)225–231

231

TheMichiganStateUniversitycampuscovers21.0km2(5200acres)andhas1.1km2(12millionft2)of atroofsurface.Ifalloftheseroofsweregreenedsimilartotheroofplat-formsinthisstudy,thenbasedonameanretentionof80.2%,theseroofswouldhaveretained377,041m3(99,603,827gal-lonsor13,315,095ft3)during2005.Ofcourse,retentiononanyroofdependsonrainfalldistributionthroughouttheyear,theintensityofeachevent,ambientairtemperatures,plantselec-tion,andthein uenceoflocalenvironmentalconditionsonevapotranspiration.

references

Barnes,K.,Morgan,J.,Roberge,M.,2001.Impervioussurfacesand

thequalityofnaturalbuiltenvironments.ReportpreparedforTowsonUniversity’sNASA/Raytheon/Synergyproject.

Bastin,L.,Thomas,C.D.,1999.Thedistributionofplantspeciesin

ndscapeEcol.14,493–507.Carter,T.L.,Rasmussen,T.C.,2006.Hydrologicbehaviorof

vegetatedroofs.J.Am.WaterRes.Assoc.42(5),1261–1274.Connelly,M.,Liu,K.,2005.GreenroofresearchinBritish

Columbia—anoverview.In:Proceedingsofthe3rdNorthAmerican:GreenRoofConferenceonGreeningRooftopsforSustainableCommunities,Washington,DC,May4–6.TheCardinalGroup,Toronto,pp.416–432.

DeNardo,J.C.,Jarrett,A.R.,Manbeck,H.B.,Beattie,D.J.,Berghage,

R.D.,2005.Stormwatermitigationandsurfacetemperaturereductionbygreenroofs.Trans.ASAE48(4),1491–1496.Deutsch,B.,Whitlow,H.,Sullivan,M.,Savineau,A.,2005.

Re-greeningWashington,DC.Agreenroofvisionbasedonenvironmentalbene tsforairqualityandstormwatermanagement.In:Proceedingsofthe3rdNorthAmerican:GreenRoofConferenceonGreeningRooftopsforSustainableCommunities,Washington,DC,May4–6.TheCardinalGroup,Toronto,pp.379–384.

Ferguson,B.K.,1998.IntroductiontoStormwater:Concept,

Purpose,Design.JohnWileyandSons,Inc.,NewYork.

Getter,K.L.,Rowe,D.B.,2006.Theroleofextensivegreenroofsin

sustainabledevelopment.HortScience41(5),1276–1285.Hawkins,R.H.,1993.Asymptoticdeterminationofrunoffcurve

numbersfromdata.J.Irrig.Drain.Eng.119(2),334–345.Liesecke,H.J.,Borgwardt,H.,1997.Abbauvonluftschadstoffen

durchextensivedachbegrunungen¨(degradationofair

pollutantsbyextensivegreenroofs).StadtundGrun

¨46(5),245–251.

Liesecke,H.J.,1998.Dasretentionsvermogen¨von

dachbegrunungen¨(waterretentioncapacityofvegetated

roofs).StadtundGrun

¨47(1),46–53.Liesecke,H.J.,1999.Extensivebegrunung¨bei5 dachneigung

(extensiveroofgreeningsona5 slope).StadtundGrun

¨48(5),337–346.

Liu,K.,2003.Engineeringperformanceofrooftopgardens

through eldevaluation.In:Proceedingsofthe18th

InternationalConventionoftheRoofConsultantsInstitute,pp.93–103.

Liu,K.,Minor,J.,2005.Performanceevaluationofanextensive

greenroof.In:Proceedingsof3rdNorthAmericanGreenRoofConference:GreeningRooftopsforSustainableCommunities,Washington,DC,May4–6.TheCardinalGroup,Toronto,pp.385–398.

Meng,Q.,Hu,W.,2005.Roofcoolingeffectwithhumidporous

medium.EnergyBuildings37,1–9.

Mentens,J.,Raes,D.,Hermy,M.,2006.Greenroofsasatoolfor

solvingtherainwaterrunoffproblemintheurbanized21stcentury?LandscapeUrbanPlan.77,217–226.

Moran,A.,Hunt,B.,Jennings,G.,2004.ANorthCarolina eld

studytoevaluategreenroofrunoffquantity,runoffquality,andplantgrowth.In:Proceedingsofthe2ndNorthAmericanGreenRoofConference:GreeningRooftopsforSustainableCommunities,Portland,OR,June2–4.TheCardinalGroup,Toronto,pp.446–460.

NRCS,2004.NationalEngineeringHandbook:Hydrology.National

SoilConservationService(FormerlySoilConservationService),USDA,Washington,DC(Chapter4).

Peck,S.W.,2005.Toronto:AModelforNorthAmerican

InfrastructureDevelopment.In:EarthPledge.GreenRoofs:EcologicalDesignandConstruction.SchifferBooks,Atglen,PA,pp.127–129.

Schade,C.,2000.Wasserruckhaltung¨undAb u beiwertebei

dunnschichtigen¨extensivbegrunungen.¨StadtundGrun¨49(2),95–100.

Scholz-Barth,K.,2001.Greenroofs:stormwatermanagement

fromthetopdown.Environ.Des.Constr.4(1),63–70.

Simmons,M.,Gardiner,B.,2007.Theeffectsofgreenroofsina

sub-tropicalsystem.GRHC.

Underwood,A.J.,1998.ExperimentsinEcology:TheirLogical

DesignandInterpretationusingAnalysisofVariance.UniversityPress,Cambridge.

USEPA(UnitedStatesEnvironmentalProtectionAgency),1994.

TheQualityofOurNation’sWater:EPAOf ceofWater,Washington,DC,EPA-841-S-94-002.

USEPA,2003.CoolingSummertimeTemperatures:Strategiesto

EPA,Washington,DC,EPA-430-F-03-014.

VanWoert,N.D.,Rowe,D.B.,Andresen,J.A.,Rugh,C.L.,Fernandez,

R.T.,Xiao,L.,2005.Greenroofstormwaterretention:effectsofroofsurface,slope,andmediadepth.J.Environ.Qual.34(3),1036–1044.

Villarreal,E.L.,Bengtsson,L.,2005.ResponseofaSedum

green-rooftoinpidualrainevents.Ecol.Eng.25(1),1–7.WaterResourcesGroup,1998.Waterlaws:understandingthe

problemsfacingurbanwatersheds,20October2005./guest/guest1.

YokTan,P.,Sia,A.,2005.Apilotgreenroofresearchprojectin

Singapore.In:Proceedingsof3rdNorthAmericanGreenRoofConference:GreeningRooftopsforSustainableCommunities,Washington,DC,May4–6.TheCardinalGroup,Toronto,pp.399–415.

本文来源：https://www.bwwdw.com/article/edfq.html