1
2016SeattleTreeCanopyAssessment
Seattlehasalong‐standingcommitmenttoitsurbantrees.Becauseofthemanysocial,environmental,andeconomicbene itsurbantreesprovide,theyareessentialtoenhancingthecommunity’squalityoflife,especiallyasSeattlegrows.
TheUrbanForestStewardshipPlan(UFSP),adoptedbyCouncilin2013,hasagoaltoreach30%treecanopycoverby2037.Theplanalsooutlinesadditionalgoalsforathrivingurbanforestthatincludeahealthydiversityoftreespeciesandagesaswellastheimportanceofcommunityparticipationincaringforandplantingtrees.MeasuringtreecanopyTreecanopycoveristhelayerofbranches,stems,andleavesoftreesthatcoverthegroundwhenviewedfromabove.CanopycoverassessmentstellustheextentofSeattle’streesandwheretheyarelocated;andhelpusinformurbanforestryworkplanning,management,andinvestments.
TherehavebeenseveraltreecanopycoverassessmentsinSeattleoverthelastdecade,eachusingvaryingmethodologiesandyieldingdifferentresults.Duetothedifferingtechnologiesandmethodologies,resultscannotbecomparedbetweenstudies.The2016LiDARassessmentTheCityofSeattleobtainedLiDAR(lightdetectionandranging)datain2016toassessprogresstowardsachievingour30%canopycovergoal.ThisstudyrepresentsthemostaccurateaccountingofSeattle’surbancanopytodateandshowsSeattlehas28%canopycover.
Theassessmentprovidesthefoundationforunderstandingthequantity,distribution,andcon igurationoftreecanopyinSeattle.Thetrue
valueofthisstudywillberealizedwhentheresultsareusedtoguideurbanforestrypolicyandmanagementeffortssuchasestablishinglocalizedcanopygoalsandtargetedplantingandconservationeffortstomaximizelimitedresources.
Thestudyfocusedonansweringseveralresearchquestionsincluding:progressachievingSeattle’s30%canopycovergoalcitywide,ineachoftheeightUFSPmanagementunits,andineachofthe27streettreemanagementunits;canopycoverlevelsinSeattleneighborhoodsandEquityandEnvironmentInitiativefocusareas;coniferoustodeciduoustreeratio;locationofthecity’slargesttreesandtreegroves;heatislandeffecthotspots;impactsfromdevelopment;andvolumeofvegetativematerialthatfallswithintheminimum10ftclearancedistancesofSCLdistributionandtransmissionsystems.Theresultswillbeanalyzedtoinformurbanforestryprioritiesandactionsmovingforward.2016LiDARstudy indingsThestudyexaminedourresearchquestionsaboutSeattle’scanopycovertohelpinformfutureurbanforestryworkandactions.Notable indingsinclude: Themajorityofoururbantreesarefoundin
twolocations:residentialareas(representing67%ofthelandwith72%ofSeattle’streecanopy),andintheright‐ofwaywhichisinterspersedthroughoutthecity(representing27%ofthelandand22%ofthecanopy).
Largertreesandtreegrovesareoftenpresentinparksbutarealsoonresidentialandinstitutionallands.
Canopyexceedstargetsindevelopedparks,naturalareas,multi‐family,andinstitutionalareas;isclosetotargetinsingle‐family,downtown,andcommercialareas;andisbelowtargetinindustrialareas.
ExecutiveSummary
2
Canopycoverdiffersacrossthecitybasedonlanduse,thepresenceofparksandnaturalareas,andsocio‐economicfactors.Censustractswherethepopulationtendstoberesidentsofcolorandpeoplewithlowerthanaverageincomealsohaveloweramountsoftreecanopy.
72%ofSeattle’streecanopyisdeciduousand28%isconiferous.Mostoftheconifersareonsingle‐familyland(52%).
Thepresenceoftreescanreduceurbanheatislandeffect(surfacetemperature),especiallyinland,andmitigateextremeheatimpacts.
6%ofSeattle’streecanopyiswithintheclearancedistanceSeattleCityLightmaintainsforsafetyandreliabilityofthepowergrid.
ImpactsofdevelopmentInanefforttounderstanddevelopmentimpactsonSeattle’scanopy,ourconsultantexaminedadditionaldatausinghistoricalGoogleEarthimagerytodoamini‐assessmentof80randomparcelsthatunderwentdevelopmentcomparingbeforeandafterimages.
Althoughnotstatisticallyvalid,thisexercisefoundparcelsinthedowntown,industrial,single‐andmulti‐familyareassawcanopycoverlosswhilecommercial,institutional,anddevelopedparkssawagainafterdevelopment(likelyaresultofretainedtreesmaturingovertime).
CanopycovertrendanalysisWehadoriginallyrequestedananalysisofcanopycoverchangeovertimecomparing2001LiDARto2016LiDAR.Unfortunately,the2001data,whilecutting‐edgeatthetime,isofsuchlowresolutionthatisunsuitableforcomparisontothe2016data.Togetthecanopycovertrendwedecidedtousethei‐TreeCanopyprotocol,asample‐basedapproachusinghistoricalimageryfromGoogleEarth,tocomparecanopycoverfrom2007,2010,and2015.ThisapproachfoundthatSeattlehadexperiencedapproximately2%canopycoverlossoverthespanoftheeightyearstudyperiod.Thismethodhasa+/‐3%marginoferrorandcanopycoverextent(coverage)isnotcomparabletothe2016LiDARassessment.ConclusionInsightsfromthisstudywillbeusedtoupdatetheUrbanForestStewardshipPlanandtorevisitpolicies,goals,managementefforts,andstrategiestogrowSeattle’surbanforest.Thetablebelowshowssomeofthestudy indingsbyUrbanForestStewardshipPlanManagementUnit.Localizedcanopygoalsandtargetedplantingandconservationeffortswillhelpmaximizelimitedresources.Nowthatwehaveareliablebaseline,itwillbeimportanttoperformanotherLiDAR‐basedtreecanopyassessmenttodeterminechangestocanopycoveroverthenext iveyears.
Management Unit # of acres Land area ( % of city)
% canopy cover % contribu on to
city’s canopy cover % conifer
contribu on
Single‐Family Residen al
29,918 56% 32% 63% 52%
Mul ‐family Residen al
5,646 11% 23% 9% 5%
Commercial/Mixed‐use
4,522 8% 14% 4% 2%
Downtown 815 1% 10% <1% 0.2%
Industrial 6,191 11% 6% 2% 1.3%
Ins tu onal 1,101 2% 25% 2% 2%
Developed Parks 2,578 4% 34% 6% 7.5%
Parks’ Natural Areas 2,356 7% 89% 14% 30%
3
UrbantreesinSeattleareanimportantassetthatprovidesmanysocial,environmental,andeconomicbene itsthatenhancethequalityoflifeforthecommunity.Seattle’sUrbanForestStewardshipPlan(UFSP),adoptedbyCouncilin2013,hasagoaltoreach30%treecanopycoverby2037.Theplanalsooutlinesadditionalgoalsforathrivingforestthatincludeahealthydiversityoftreespeciesandages.
Canopycoverassessmentsareanimportanttoolfortrackingprogress.ForSeattletoeffectivelymanageourtreecanopyandenactpolicies,plans,andinitiativestohelpensurearobusturbanforestforgenerationstocome,acomprehensiveunderstandingofourtreecanopyisvital.TheSeattleOf iceofSustainability&Environment(OSE),commissionedthisstudytotheUniversityofVermontSpatialAnalysisLabwhichincludessomeoftheworld’sforemostexpertsinurbantreecanopymapping.AlthoughtreecanopyforSeattlehasbeenmappedbefore,the2016LiDARtreecanopyassessmentrepresentsthemostaccurateaccountingoftreecanopytodate.
Manyofthebene itsurbantreesprovidearerelatedtothesizeandstructureofthetreecanopy,whichisthelayerofbranches,stems,andleavesoftreesthatcoverthegroundwhenviewedfromabove.
Understandingtreecanopyisanimportantstepinurbanforestryplanning.Atreecanopyassessmentprovidesanestimateoftheamountoftreecanopycurrentlypresentinacommunity,wherethecanopyislocated,andtheamountoftreecanopythatcouldtheoreticallybeestablished.Thetreecanopyassessmentcanbeusedbyabroadrangeofstakeholderstohelpcommunitiesplanagreenerfuture.
OSEinpartnershipwiththeUrbanForestryTeammembers(SeattleCityLight,SeattleDepartmentofConstructionandInspections,SeattleDepartmentofTransportation,SeattleParksandRecreation,andSeattlePublicUtilities)ledtheefforttoobtain2016LiDARdatatoperformacanopycoverassessmentfortheCityofSeattleinsupportoftheUFSPgoals.AtOSE’srequest,theSpatialAnalysisLaboratoryanalyzedthedatatorespondtoprojectresearchquestionsusingtheUSDAForestServices'TreeCanopyAssessmentprotocols.
Introduction
WhyisTreeCanopyImportant?
Figure 1. Study area for this project, which is the full City of Seattle boundary.
Figure 2. Tree Canopy derived from high-resolution imagery and LiDAR.
ProjectPartners
4
TreecanopywasassessedusingtheUSForestServiceUrbanTreeCanopy(UTC)Assessmentprotocols.Twodistinct,complementaryremotelysenseddatasetswereusedtomapSeattle’scanopy:LiDAR(lightdetectionandranging),acquiredunderleaf‐offconditionsin2016(Figure3A)and2015leaf‐onimagery,obtainedthroughtheUSDANationalAgriculturalImageryProgram(NAIP)(Figures3Band3C).WhileslightlylessaccuratethanLiDAR,thisimageryhasthebene itofprovidingcolorinformationthatisusefulinseparatingvegetatedfromnon‐vegetatedsurfaces.Treecanopymappingwasaccomplishedusingasemi‐automatedapproachinwhichtrees’featureswereautomaticallyextracted,thenmanuallyreviewedandeditedatascaleof1:2500.Anyobservableerrorswerecorrectedandincorporatedintothe inalresults.TheUTCAssessmentprotocolsaretheacceptedstandardformappingtreecanopyandhavebeenappliedbytheUniversityofVermontSpatialLabtoover80communitiesinNorthAmerica.Aformalaccuracyassessmentwasnotperformedforthestudy,butsimilarstudieshaveshowstohaveamarginoferrorof+/‐1%.
This2016treecanopyassessmentrepresentsthemostaccurateaccountingoftreecanopyeverdoneforSeattle,withtreesassmallaseightfeetinheightmapped(Figure3D).Datasetsaresuitableforsummarizingtheareaandpercentareaoftreecanopydowntotheindividualpropertyparcellevel.The2016LiDARdatasetwillserveasthefoundationfortrackingtreecanopychangesovertimefortheCityofSeattle.
HowwastheTreeCanopyMapped?
Figure 3. Example of the input data used to extract tree canopy data. LiDAR is on the left, true color imagery, CIR Imagery, and then tree canopy data on the right.
A B C D
Previous projects have mapped Sea le’s
tree canopy, but as this project used a
combina on of superior source data and
methodologies, any comparisons between
the various studies are not valid. Of
par cular note is the fact that prior canopy
cover es mates should not be used to
draw conclusions with respect to changes
in tree canopy over me.
5
TreecanopymetricsaregeneratedbytheUSDAForestService’streecanopymetricstool(Figure4).ThistooltakespropertyparcelsandCensustractsandcomputestheamountofExistingTreeCanopyandPossibleTreeCanopyineachgeography.ExistingTreeCanopyistheamountoftreecanopythatcurrentlyexistsbasedon2016conditions.Itiscomputedbydividingtheamountoftreecanopyineachgeographicunitbythelandarea.PossibleTreeCanopyistheamountoflandthatistheoreticallyavailablefortheestablishmentofnewtreecanopy.Itisimportanttonotethatjustbecauselandisavailableforestablishingnewtreecanopyitdoesnotmeanthatitwouldbeeithersociallydesirablenor inanciallyfeasibletoestablishtreecanopyonsuchareas.Figure5showsanexampleoftheTreeCanopymetricscomputedatthepropertyparcellevel.Theparcelboundaries(top,blacklines)representthegeographicunits.TheExistingTreeCanopy(middle)istherelativepercentageoftreecanopyineachparcel.PossibleTreeCanopy(bottom)indicatestherelativepercentageoflandineachparcelthatistheoreticallyavailablefortheestablishmentofnewtreecanopy.
HowisTreeCanopyCalculated?
Figure 5. Parcel-based tree canopy metrics.
Parcels
PossibleTreeCanopy
ExistingTreeCanopy
Figure 4. Graphical user interface for the tree canopy metrics tool. The tree canopy metrics tool is an ArcGIS-based geoprocessing model that summarizes tree canopy information based on the input polygon boundaries
6
WhatisourProgressonAchieving30%TreeCanopyCover?
Methods Results
Figure 6. Tree Canopy derived from high-resolution imagery and LiDAR. Figure 7. Tree Canopy at the city-scale derived from high-resolution imagery and LiDAR
Treecanopyiscomputedbymappingalllandcoverthendividingtheamountoftreecanopybytheamountofland.Excludingwatergivesamoremeaningfulpercentagethatcanbecomparedtoothercities.
28%ofSeattle’slandiscoveredbytreecanopy91,776‐38,271=53,505thisisthetotallandareawithoutwater15,167/53,505=28%dividingthetreecanopyareaintothelandareagivesusthepercenttreecanopycover
91,776acresTotalareaofthecity(includeswater)38,271acresAreaofwaterwithinthecityboundary15,167acresTotalareaoftreecanopy
7
Becauseofthedifferencesbetweendevelopedproperty,streetscapes,parklands,remnantforests,andotherareas,Seattle’surbanforestcannotbeviewedasasingleunitformanagementpurposes.TheUFSPde ineseightmanagementunits(MUs)thatcoverallthelandtypesinthecity.Usingtheseland‐usetypesallowsforeasycoordinationofGISmappinglayersandforrelatedplanninginitiatives.Theunitsincludeeightdistinctareasthatwereselectedbaseduponphysicalcharacteristics.AninthMU,theright‐of‐way,overlaystheeightunitssoisnotaseparatelandarea.Canopycoverisamajorindicatorofthebreadthoftheurbanforestanditsoverallhealthandvitality.TodeterminethecanopycoverofeachMU,theirindividuallandareawasdeterminedandthetreecanopypercentineachMUwascalculated.TreecanopywascomputedbothintermsoftotalareaandasapercentageofthelandareawithineachMU.
Table1identi iesthepercentageofthecity’slandmass,currentcanopycover,targetsbyMU,aswellaseachMU’scontributiontothecity’soverallcanopycover.Figure8showsMUscontributionto
overallcanopycover.CanopycoverexceedstargetsinDevelopedParks,Parks’NaturalAreas,Multif‐familyResidential,andInstitutionalmanagementunits;isclosetotargetintheSingle‐familyResidential,Downtown,andCommercial/Mixed‐usemanagementunits;andisbelowtargetintheIndustrialmanagementunit.Themajorityofoururbantreesresideintwolocations:residentialareas(representing67%ofthelandandhousing72%ofSeattle’streecanopy),andintheRight‐of‐way(ROW),whichrepresents27%ofthelandthatisinterspersedthroughoutallManagementUnits,andhouses22%ofthecanopy.
WhatisourProgressonAchievingCanopyCoverTargetsontheUrbanForestStewardshipPlanManagementUnits?
Methods
Results
Management Unit Land area ( % of city)
2016 canopy cover
2037 canopy goal (set in 2007)
% contribu on to city’s canopy cover
Single‐Family Residen al 56% 32% 33% 63%
Mul ‐family Residen al 11% 23% 20% 9%
Commercial/Mixed‐use 8% 14% 15% 4%
Downtown 1% 10% 12% <1%
Industrial 11% 6% 10% 2%
Ins tu onal 2% 25% 20% 2%
Developed Parks 4% 34% 25% 6%
Parks’ Natural Areas 7% 89% 80% 14%
City total 100% 28% 30% 100%
Right‐of‐way (runs through all other MUs)
27% 23% 24% 22%
Figure 8. Contribution to the city’s overall canopy cover by Management Unit.
Table 1. Percentage of city’s landmass,, current canopy cover, targets by MU and MU contribution to city’s canopy cover
8
TheamountofExistingTreeCanopywascomputedforthirteenneighborhoods,whoseboundariesdonotpreciselyalignwiththeactuallandarea.Thisresultsinthe“water”neighborhoodhavingasmallamountoftreecanopyduetothepresenceoftreesalongthewater’sedge.
Theamountoftreecanopyineachareaisre lectiveoflanduseandurbanization,withindustrialareasandtheurbancorehavingtheleastamountoftreecanopy(Figure9).Downtownhas10%canopycover,mostlyintheROW.GreaterDuwamish,themostheavilyindustrialarea,has15%canopycover.Delridge,withitslargepatchesofconnectedtreecanopyisabovethecityaveragewith38%canopycover.TheNorthneighborhoodhasarobustamountoftreecanopywithinitsresidentialareas(35%).TheCentral,East,Northeast,Northwest,Southeast,andSouthwestneighborhoodsallhaveatleast30%canopycover.Thiscanlargelybetracedtorobusttreecanopyofresidentiallandscoupledwithlargerexpansesoftreecanopyinparks,theROW,andotherprotectedareas(Figure10).
WhatistheCanopyCoverinDifferentAreasofSeattle?
Methods
Results
Figure 9. Map of existing tree canopy for each of Seattle’s neighborhoods.
Figure 10 Percent of neighborhoods’ land that is covered by tree canopy.
9
TheSeattleDepartmentofTransportation(SDOT)isinventoryingstreettreesaspartoftheStreetTreeManagementPlan.SDOTdividedthecityinto27StreetTreeManagementUnits.Havingcleargeographicboundariessupportsestablishmentandtrackingofperformancemeasures.Inventorydatawillinformtreemaintenanceandreplacementeffortsandwillhelpbalanceresourcesmoreequitably.
CanopycoverwascalculatedforeachunitandfurthersubdividedintowhetherornotitfellwithintheROW.Itisimportanttonotethata
limitationofLiDARanalysisisthatthelocationofthetreestemcanbeobscuredduetothetopdownnatureofthetechnology,makingtheanalysisoftreeswithintheROWlessprecise.
TheamountoftreecanopyineachSDOTStreetTreeManagementUnit,alongwiththeproportionofthetreecanopythatfallswithintheROW,differsubstantiallyfromunittounit(Figures11and12).Unit18hasthehighestpercentoftreecanopyat39%,with15%beingintheROW.Atthelowend,Unit27hasthelowesttreecanopyat10%withthe70%ofthecanopybeingintheROW.
WhataretheCanopyCoverlevelsinSDOT’sStreetTreeManagementUnitsandtheRight‐of‐Way?
Methods
Results
Figure 11. Percent of existing tree canopy for each of the SDOT’s 27 street tree management units.
Figure 12. Percent of tree canopy as ROW or Not ROW in relation to each of the SDOT’s 27 street tree management units.
10
In2015,MayorEdMurraylaunchedtheEquityandEnvironmentInitiative(EEI)todeepenSeattle'scommitmenttoraceandsocialjusticeinenvironmentalworkalsocreatingtheEnvironmentalActionAgenda(EEA).TheEEAisaseriesofcommunity‐developedgoalsandstrategiesthataddressenvironmentalinequitiesandcreateopportunitiesforcommunitiesofcolor,refugees,peoplewithlowincomes,andlimitedEnglishpro iciencyindividualstobecomeleadersinSeattle'senvironmentalmovement.Twoenvironmentalequitymeasureswereselectedforanalysis:peopleofcolorandpeoplewithin200%ofthepovertylevel.Thesetwofactorswereanalyzed,mapped,andgraphedalongsidetreecanopy.Itshouldbenotedthatalthoughdemographicfactorsmaybecorrelatedwithtreecanopy,thesecorrelationsdonotnecessarilyequalcausation.
Thereisastatisticallysigni icantinverserelationshipbetweentreecanopyandbothpeopleofcolorandpeoplewithin200%ofthepovertylevel.TheanalysisfoundthatinCensustractswithloweramountsoftreecanopymoreofthepopulationtendstobepeopleofcolorandhavelowerincomes.Incensustractswithhighnumbersofpeopleofcolor,treecanopyisaslowas11%whileinareaswithnotmanypeopleofcolorthereis55%canopycover(Figure14).Itisimportanttonotethatalthoughthereisageneralinverserelationship,therearenumerousexceptions.SomelocationswithinSeattlethathavethehighestconcentrationsofpeopleofcolorandresidentsunder200%ofthepovertylevelhavearelativelyhighpercentageoftreecanopyduetothepresenceofparksandstreettrees.
WhatistheRelationshipBetweenTreeCanopyandEnvironmentalEquity?
Methods Results
Figure 13. Map of percent people of color and percent tree canopy for each Census tract.
Figure 14. Figure describing percent tree canopy in relation to people of color. Each dot represents an EEA polygon.
The analysis found that in Census
tracts with lower amounts of tree
canopy more of the popula on
tends to be people of color and
have lower incomes.
11
Coniferoustreesprovidegreaterenvironmentalbene itsbecausetheytendtomaintaintheircanopyyear‐round,includingduringtherainyseason,helpingslowdownandreducestormwaterrun‐off.Theyalsoabsorbmorecarbondioxideandairpollutants.Achallengeconiferspresentisthatduetotheirlongerlifespans,theytendtoachievegreatersizerendingthemnotviableinmanylocations.Homeownersareoftenreluctanttoplantconifersintheiryardsandtheyarenotalwaysagood itforstreettrees.Tocalculatethedeciduoustoconiferoustreeratio,treecanopywasseparatedintoapproximateindividualtreecanopiesusingatreecanopysegmentationroutinedevelopedforthisproject.Eachtreecanopyshapewasthenclassi iedaseitherdeciduousorconiferousbasedonitsmorphologicalpropertiesintheLiDARdata.
Coniferstendtobeconicalinshape,whiledeciduousspecies,morerounded.TheoptimalmethodofseparatingoutevergreenspeciesistouseacombinationofLiDARandleaf‐offimagery,butnosuchimagerywasavailabletobeusedintheproject.Theprojectteamdecidedtoaccountforconiferousasaproxyforevergreenspecies.Thedeciduous/coniferousbreakdownwasthensummarizedforeachManagementUnitandthenbasedonwhetherornotthecanopywasinsideoroutsideoftheROW.
Citywide,72%oftreesaredeciduousand28%areconiferous.17%oftheconifersareintheROW.Parks’NaturalAreasistheonlymanagementunitwithamajorityofconifers(60%).Downtownhadthelowestpercentofconiferoustreesatonly10%,whichdemonstratesthechallengewefaceaccommodatingconifersintheright‐of‐way.
WhatistheDeciduoustoConiferousTreeRatioCitywide,byManagementUnit,andtheRight‐of‐Way?
Methods
Results
Figure 15. Percent of deciduous to coniferous tree ratio by each UFSP Management Unit
12
Theprojectteamde inedlargetreesandtreegrovesbasedondiameteratbreastheight(DBH).Forthepurposeofthisstudy,largetreesmusthaveatleasta30‐inchDBHandtreegroveshavetocontainaminimumofeighttrees,eachwitha12‐inchminimumDBHandformingacontiguouscanopy(excludingstreettrees).AsDBHcannotbedirectlymeasuredfromoverheadremotesensingdata,weusedheightinformationfromtheLiDARdatasetasaproxybasedonequationsdevelopedbyJenkinsetal(2003).
Seattleisestimatedtohave6,338largetrees(Figure17)and3,188treegroves(Figure18).Notsurprisingly,thelargetreesandtreegrovesareoftenpresentinparksandotherforestedareasthroughoutthecity.Althoughthelargetreesandtreegrovesarelargelyabsentfromthemosturbanizedareassomedoexistonresidentialandinstitutionallands.
WherearetheLargestTreesandTreeGrovesLocated?
Methods Results
Figure 16. Locations of the largest trees at the city scale. Figure 17. Locations of tree groves at the city scale.
13
Mappingtreecanopychangeovertimerequiresconsistent,high‐qualitydata,withoutwhichitisnotpossibletodetermineiftheresultingtrendisduetoactualcanopychangeorduetoanomaliesinthesourcedata.Theoriginalstudyscopeincludedcomparing2001LiDARwith2016LiDARtodeterminecanopycoverchangeovertime.Unfortunately,the2001dataset,whileconsideredcutting‐edgetechnologyatthetime,isofverylowresolutionwhichpreventedreliablecomparisontothenew,higher‐qualitydatafrom2016. However,in2015,OSEcommissionedananalysisusingUSDAForestServicei‐TreeCanopyprotocols,whichconsistofasample‐basedapproachthatestimatescanopybydeterminingthepresence/absenceoftreecanopyatagivenlocationusingaerialorsatelliteimagery.Theamountofchangewasdeterminedthroughthemanualinterpretationof4,000strati iedrandompointsgeneratedforthreeyears:2007,2010and2015.
The2015analysisestimateda2%lossoverthespanoftheeight‐yearstudyperiodwitha+/‐3%marginoferror(Figure18).The2016LiDARdatasetwillserveasthefoundationfortrackingtreecanopychangesovertimefortheCityofSeattlemovingforward.Anumberoffactorscouldcausedifferencesincanopyestimatesincludinglimitationsofusingasample‐basedapproachandthathistoricalimageryfromGoogleEarthisnotcollectedatthesametimeofday,causingshiftsoftreecanopylocation.Treecanopylosscouldinitiallybecausedbyincreaseofimpervioussurfaceduetodevelopment.Naturalgrowthofmaturetreesandtreeplantingstendtooffsetcanopyloss.
HowhasSeattle’sCanopyCoverChangedsince2007?
Methods Results
Thedifferencesinthe2015approachandthe2016LiDARmakesthestudies’resultsnotdirectlycomparabletoone
another.Priorestimatesoftreecanopyshouldnotbeusedto
drawconclusionswithrespecttochangesintreecanopyovertime.
Figure 18. Percent tree canopy cover for each UFSP Management Unit in 2007, 2010 and 2015. The gray area represents the standard error.
14
Developmentimpactontreecanopywasexploredbyrandomlyselecting10developmentpointsfromeachUFSPManagementUnittotaling80points.Using2007‐2017historicalimageryfromGoogleEarth,treecanopywasmappedbyparcelforbeforeandafterdevelopment(Figure19).Percenttreecanopywascalculatedbydividingthetotalareaoftreecanopybythetotalareaoftheparcels.
Althoughthispieceisnotstatisticallyvalid,itisadetailedanalysisofhowdevelopmentimpactedtreecanopyateachidenti iedpointandprovidesgeneralinsightsintohowdevelopmentcanimpactcanopycoverovertime..DowntownandSingleFamilyhadthemosttreecanopyloss.(Figure20).
WhatImpactcanDevelopmentHaveonTreeCanopy?
Results
Figure 19. Example of the impacts of development in a single family management unit parcel. The left image is from 2011 and the right is 2016.
Methods
Figure 20. Percent tree canopy cover for each management unit before and after development occurred.
Land Area (% of city)
Single Family 2011 Single Family 2016
15
Theurbanheatislandeffectisproducedbydenseconcentrationsofbuildings,pavement,andothersurfacesthatabsorbandretainheat.Thisincreasesairpollution,costsrelatedtoairconditioning,andheat‐relatedhealthconditionsdisproportionallyimpactingvulnerablepopulations.Treecanopyhelpsreduceheatislandeffectmitigatingimpacts.
SurfacetemperatureobtainedfromLandsat8satellitethermalimagerycollectedonSeptember20,2016wasusedtoestimatetheurbanheatislandeffect.Landsat8sensesthermalenergyataresolutionof60‐meters.Surfacetemperaturewasintegratedwiththe2016LiDARtreecanopydatatoanalyzehowthepresenceoftreecanopyaffectstheurbanheatisland.
DeterminingtheurbanheatislandinSeattleischallenginggiventhatmuchofthecityissurroundedbywater,whichdampenstheurbanheatislandeffect.Nevertheless,intheinlandareastreesclearlyhelpreducesurfacetemperature,therebyreducingtheurbanheatisland.ThisstudyproducedGISshape ilesforexistingtreecanopy,thermalimagery,andpossibletreecanopy(Figure21)thatcanbeusedtoidentifytreeplantinglocationstominimizetheimpactsofurbanheatislandhotspotsbyreducingsurfacetemperatures,aswellastoprioritizetreeplantingeffortstomitigateequityissues.
HowdoesTreeCanopyReducetheUrbanHeatIslandEffect?
Methods Results
Figure 21. Seattle’s urban heat island hot spots correspond to low tree canopy areas..
16
SeattleCityLightmaintainsvegetationclearances(pruningtreesawayfrompowerlines)forsafetyandelectricalservicesystemreliability.SCLmaintainsclearancesinaccordancewithutilitybestmanagementpracticesandtree‐industrystandards.ToisolatecanopythatintersectsSCL’selectricalinfrastructure,a10‐footbufferwascreatedaroundallprimaryandsecondarydistributionlines,transmissionsupportstructures,and115kV‐
230kVoverheadwires(Figure22).BiomassequationsfromJenkinsetal.(2003),wereusedtocomputevolume.Thebiomassequationsrequiresaconsiderationoftreediameteratbreastheight(DBH)andspecies.AsDBHwasnotavailable,heightwasusedasproxy.Averageswereobtainedforthemostcommonconiferousanddeciduousspecies:blackcottonwood,redalder,westernredcedar,DouglasFir,bigleafmaple,Scouler’swillow,westernhemlock,andbeakedhazelnut.Densitieswere540kg/m3and470kg/m3fordeciduousandconiferoustrees,
respectively.Diameter‐at‐breast‐heighttotreeheightmodels(Figure23)wascreatedbycombiningminimum,maximum,andmeandataforeachparameterfromseveraljournals(FierkeandKauffman,2005;Garmanetal.,1995;andHanusetal.1999).
Thevolumeoftreecanopythatintersectstheminimum10‐footclearancedistancesforSeattleCityLight’sdistributionandtransmissionsystemsis1,328,000ft3or6%ofSeattle’stotaltreecanopyvolume(Table2).SCLwillusethisinformationandadditionalanalysisthatisbeingperformedoutsidethescopeofthisstudytoinformvegetationmanagementefforts.
WhatistheVolumeofTreeCanopythatIntersectstheMinimum10ftClearanceDistancesforSCL’sElectricalSystem?
Methods
Results
Figure 22. Coniferous and deciduous trees intersecting the 10ft minimum distance of SCL infrastructure.
Figure 23. DBH by height regression to produce a simple polynomial to best fit the model. This model incorporates data of all species before coniferous/deciduous differentiation.
Table 2. Volume (cubic feet) of Seattle trees and trees that intersect within 10ft of SCL infrastructure.
17
Treecanopy inSeattle isavitalasset thatprovidesmultipleecosystemservicessuchasstormwaterrunoffreduction,improvedairquality,decreasedcarbonfootprint,enhancedqualityoflife,savingson energybills,andhabitatforwildlife.
The2016LiDARassessmentrepresentsthemostaccurate,comprehensive,anddetailedaccountingofSeattle’scanopycovertodate.Itprovidesthefoundationforunderstandingthequantity,distribution,andcon igurationoftreecanopywithinSeattleandestablishesthebaselineforfuturecanopycoverchangeanalysis.
ThemajorityofSeattle’surbantreesarefoundintwolocations:residentialareas(representing67%ofthelandwith72%ofSeattle’streecanopy),andintheright‐of‐way,interspersedthroughoutthecity(representing27%ofthelandwith22%ofthecanopy).
Largertreesandtreegrovesareoftenpresentinparksbutarealsoonresidentialandinstitutionallands.
Canopyexceedstargetsindevelopedparks,naturalareas,multi‐family,andinstitutionalareas;isclosetotargetinsingle‐family,downtown,andcommercialareas;andisbelowtargetinindustrialareas.
Canopycoverdiffersacrossthecitybasedonlanduse,thepresenceofparksandnaturalareas,andsocio‐economicfactors.Censustractswherethepopulationtendstoberesidentsofcolorandpeoplewithlowerthanaverageincomealsohaveloweramountsoftreecanopy.
72%ofSeattle’streecanopyisdeciduousand28%isconiferous.Mostoftheconifersareonsingle‐familyareas.
Thepresenceoftreescanreduceurbanheatisland(surfacetemperature),especiallyinland.
6%ofSeattle’streecanopyiswithintheclearancedistanceSeattleCityLightmaintainsforsafetyandreliabilityofthepowergrid.
Althoughnotstatisticallyvalid,amini‐assessmentof80randomparcelsfoundthatdevelopmentledtocanopycoverloss.
Aseparatesample‐basedsatelliteimagerytrendanalysisdonein2015suggestsanoverallcanopycoverlossofaround2%from2010to2015,withamarginoferrorof+/‐3%.
1. Thevalueofthisstudywillberealizedwhentheresultsareusedtoguideurbanforestrypolicyandmanagementefforts.
2. Preservingtreecanopyisasimportantasnewtreeplantinginitiatives.Establishedmaturetreegrowthcontributestocanopycovergain.
3. Studydatacanbeusedtoestablishlocalizedcanopygoalsandtargetedplantingsandconservationeffortstomaximizelimitedresources.Selectingaspeci icbene ittobuildanengagementcampaigncanincreasethesuccessintreeplantingandconservationactions,particularlywhenanaudienceisalreadygalvanizedaroundaparticularissue(e.g.engagingresidentsconcernedaboutairqualityissuesinaspeci icneighborhoodintreeplantingeffortsinthatarea).
4. ItisrecommendthattheCityofSeattleupdateitsUrbanForestryStewardshipPlanwithinformationderivedfromthisstudy.Theinformationcanbeusedtorevisitexistinggoalsandtargets.
5. AnotherLiDAR‐basedassessmentshouldbeplannedtodeterminechangestothetreecanopyinSeattlewithinthenext iveyears.Suchassessmentcanprovideinformationonhoweffectivetreeplantingandpreservationeffortshavebeen,inadditiontounderstandinghowotherfactors(e.g.development,drought,pests,etc.)maybeimpactingtreecanopy.Futureassessmentswillonlybemadepossibleifcontinuedinvestmentsinhigh‐resolutionremoteimageryandLiDARdataacquisitionaremade.UndertakingLiDAR‐basedassessmentsinthefuturewillallowfortrendanalysiswithcomparabletreecanopydatatobemade.
Conclusions
Study indings
Usingthe indings
18
Additional Information Preparedby:JarlathO’Neil‐DunneUniversityofVermontSpatialAnalysisLaboratoryjoneildu@uvm.edu802.656.3324
Formore info on theUrban Tree Canopy Assessmentpleasevisithttp://nrs.fs.fed.us/urban/UTC/
TreeCanopyAssessmentTeam:NoahAhles,LukeBan,JarrettBarbuto,NoahBell,JasonBlack,ErnieBuford,EmmaButter ield,JosePabloBrenesCoto,JacobCiof i,KristineCorey,KaiDarke,TaylerEngel,EmmaEstabrook,NathanielFuchs,MikeFranck,LindseyFreitag,DanGordon,JacobKing,SeanMacFaden,JaredMaher,ElizabethMcElwee,OwenMoseley,JarlathO’Neil‐Dunne,AnnaRoyar,KellySchulze,MatthiasSirch,ConnorSullivan,PatrickSullivan,andJamesRambone.
Figure 25. 2016 LiDAR for Seattle. The acquisition of such data in the future will be crucial for tracking tree canopy change over time.
Fierke,M.K.andJ.B.Kauffman.2005.Structuraldynamicsofriparianforestsalongablackcottonwoodsuccessionalgradient.Garman,S.L.,Acker,S.A.,Ohmann,J.L.,andT.A.Spies.1995.Asymptoticheight‐diameterequationsfortwenty‐fourtreespeciesinWest‐
ernOregon.ResearchContribution,10:5‐22.Hanus,M.L.,Marshall,D.D.,andD.W.Hann.1999.Height‐diameterequationsforsixspeciesintheCoastalRegionsofthePaci icNorth‐
west.ResearchContribution,25:5‐11.Jenkins,J.C.,Chojnacky,D.C.,Heath,L.S.andR.A.Birdsey.2003.National‐ScaleBiomassEstimatorsforUnitedStatesTreeSpecies.Forest
Science,49:12‐35.
Figure 24. Seattle canopy cover by Urban Forest Stewardship Plan management unit..
2016 Sea le Canopy Cover