ISSN 1911-5814 17 Prairie Perspectives: Geographical Essays 2018, 20: 17–25 An analytical comparison of flood zones derived from point cloud LiDAR data and historical flood data: A case study of Moose Jaw, Saskatchewan, Canada Yulu Peng Department of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, Moose Jaw Campus, Saskatchewan Abdul Raouf Department of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, Moose Jaw Campus, Saskatchewan Muhammad Almas Water Security Agency, Moose Jaw, Saskatchewan The high accuracies of point cloud data captured using light detection and ranging (LiDAR) offer an advantage over traditional surveying techniques that are used to extract elevation information in developing a digital elevation model (DEM) of an inacces- sible area. This research demonstrates the use of point cloud LiDAR data for flood zone mapping in an urban environment, and specifically in Moose Jaw, Saskatchewan. Base flood signatures around water bodies were extracted from high resolution aerial photographs to establish base flood elevation (BFE) using 0.25 m elevation contours derived from LiDAR data. A surface eleva- tion of 0.5 m above the BFE was classified as a flood zone. It was estimated that 787 ha of lands within Moose Jaw’s boundar- ies fall within the flood zone. The developed flood zone was found to be 184 ha smaller than one developed using 500-year historical flood recurrence data. Spatial analysis techniques were then used to identify historical inaccuracies and changes in land cover as the possible causes for these changes. The study demonstrated the need for regular mapping of flood zones as they can change over time. Remote sensing technologies utilizing high resolution aerial photographs and point cloud LiDAR data can be used effectively for this purpose. Keywords: LiDAR, DEM, flood zones, remote sensing, GIS This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited. Correspondence to: Abdul Raouf, Department of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, PO Box 1420, Moose Jaw, SK S6H 4R4 Email: [email protected]Key Messages • Regularly updated and accurate flood zone maps should be made publicly available to increase awareness of risks to per- sonal property. • Existing flood zone maps are based on historical flood recurrence data and are outdated as flood zones can change. • The latest remote sensing technologies, such as aerial images and LiDAR, can be used to update existing flood zone maps and to estimate the volume of flooded waters.
An analytical comparison of flood zones derived from point cloud LiDAR data and historical flood data: A case study of Moose Jaw, Saskatchewan, Canada
Yulu PengDepartment of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, Moose Jaw Campus, Saskatchewan
Abdul RaoufDepartment of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, Moose Jaw Campus, Saskatchewan
Muhammad AlmasWater Security Agency, Moose Jaw, Saskatchewan
The high accuracies of point cloud data captured using light detection and ranging (LiDAR) offer an advantage over traditional surveying techniques that are used to extract elevation information in developing a digital elevation model (DEM) of an inacces-sible area. This research demonstrates the use of point cloud LiDAR data for flood zone mapping in an urban environment, and specifically in Moose Jaw, Saskatchewan. Base flood signatures around water bodies were extracted from high resolution aerial photographs to establish base flood elevation (BFE) using 0.25 m elevation contours derived from LiDAR data. A surface eleva-tion of 0.5 m above the BFE was classified as a flood zone. It was estimated that 787 ha of lands within Moose Jaw’s boundar-ies fall within the flood zone. The developed flood zone was found to be 184 ha smaller than one developed using 500-year historical flood recurrence data. Spatial analysis techniques were then used to identify historical inaccuracies and changes in land cover as the possible causes for these changes. The study demonstrated the need for regular mapping of flood zones as they can change over time. Remote sensing technologies utilizing high resolution aerial photographs and point cloud LiDAR data can be used effectively for this purpose. Keywords: LiDAR, DEM, flood zones, remote sensing, GIS
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution, andreproduction in any medium, provided the original work is properly cited.
Correspondence to: Abdul Raouf, Department of Geomatics and Surveying Engineering Technology, Saskatchewan Polytechnic, PO Box 1420, Moose Jaw, SK S6H 4R4 Email: [email protected]
Floods are an important natural process and one of themostcommon and expensive natural hazards inCanada,where re-latedlossestopersonalpropertyareestimatedtocostupto$600millionperyear(Thistlethwaiteetal.2017).Thecostoffloodre-latedhazardsisincreasingduetoclimatechangeandcontinuousdevelopment inflood-proneareas.Poormaintenanceof stormwaterandfloodprotectioninfrastructuremayaddtotheriskofflooding (Thistlethwaite et al. 2017).Flood relateddamage isnotlimitedtodirectfinancialimpacts,asarangeofsocial(e.g.,business failure, unemployment and population displacement)and public health (e.g.,mold-related health effects, post-trau-maticstressdisorder,depression,andanxiety)relatedissuesalsofollowfloods(Levinetal.2007;Lamondetal.2015).Frequentoccurrences of severe storms due to thewarming atmosphereandslowingjetstreamshascreatedanurgentneedforanaware-ness campaign, through informationdissemination, forproperprotection andmanagement of floods. Homeowners can onlyplaytheirpartinreducingfloodrisksiftheyareawareoftheop-tionsavailabletothem.However,themajorityofthepopulationlivinginfloodproneareasisunawareofitsvulnerability.Ana-tionalsurveyconductedbytheUniversityofWaterloorevealedthat only 6% of participants (2300 home owners) knew thattheir property or homewas locatedwithin a designatedhigh-riskfloodzone, and89%ofhomeownersdidnothavefloodinsurance for their property (Thistlethwaite et al. 2017).Out-datedfloodmapsorunavailabilityofupdatedmapscanbeoneofthemanycausesforthislackofinformation.FloodDamageReduction(FDR),startedin1975,wasaCanadiangovernmentinitiativetoproducehighqualityfloodriskmapsforallurbanmunicipalitiesinthecountry.Theprogramwasalsointendedtocoordinatefederalandprovincialstrategiestodiscouragefuturedevelopment in flood zones. Zoning authorities were encour-aged to zone on the basis of flood risks. Flood-related finan-cialassessmentandmanagement isadifficult task forseveralreasonsincludingpopulationgrowth,assetvalueincrease,andthevulnerabilityofinfrastructurewithinfloodproneareas.Inacontinuingefforttominimizetheeffectsofflooding,theGov-ernment ofSaskatchewan introduced theDevelopmentAct of2007.TheActrequiresthatallmunicipalitiesinSaskatchewanmustidentifyareaspronetonaturaldisasterbeforeinitiatinganyplanning,rezoning,anddevelopmentactivitieswithinmunicipalboundaries(GovernmentofSaskatchewan2007).Mapsoftheseareasneedtobeupdatedonaregularbasistoincludethelatestinformationasfloodzoneschangeovertime.
Risk assessment, analysis, andmapping of areas prone tonaturaldisasterscomprisesacomplexsetofprocesses.Itmayinvolvea combinationofdifferentgeomatics technologies in-cluding remote sensing and geographic information systems(GIS).SomeofthecommonreasonsforfloodinginSaskatch-ewanincludesnowmeltrunoff,ice/snowjams,extremerainfall,and structural failure (Sandink et al. 2010). Identification andcontinuousmonitoringofpossible snow jamareas canplay acrucial role infloodprevention.Thesemay includenaturalorartificial obstructions towaterflow including riverbends, de-
creasedriverslopes,tributarymouths,dams,andbridges.Waterheldbackduetoice/snowjamscancauseflashflooding.Thoughfreeze-up ice/snow jamsoccurring in earlywinter are usuallyless damaging, break-ups occurring in early spring can causesignificant damage to infrastructure and contribute to springflooding in low lyingareas.Spring thermalprocessesdeterio-rateicecovercausingittofractureintolargefloatingiceslabswhichcanaccumulateintoicejams.Kineticenergygeneratedfromfloating ice can damage infrastructure and erode streambanks.Additionally,rainthatfallsontopofsnoworsnowmeltduetoincreasingspringtemperaturescantriggersnowbreakupandcauseicejams.Thesoilunderneathathicksnowpackofalongprairiewinter is saturated and cannot absorbwater fromsnowmeltorrain,addingitdirectlytotheriversandstreams.Icejamscancreaterapidincreasesinupstreamwaterlevelsandconsequentflooding.Additionally,suddencollapsesoficejamscanviolently increasedownstreamwater levelsandwaterve-locities(Scrimgeouretal.1994).Theaverageannualcostasso-ciatedwithicejamrelateddamageinCanadaisapproximately$20million(Lawfordetal.1995).
Heavysummerrainsareanothercommoncauseoffloodingin areaswhere a large percentage of the surface is imperme-able,orwhenprolongedwetperiodshavesaturatedtheground.Soilcompression,pavedroads, impermeableparkinglots,androoftopsresultingfromurbanizationcangreatlyincreasewaterrunoffcausedbyprecipitation,resultinginurbanfloods(Law-fordetal.1995).Heavyrainfalloverashortperiodoftimecansignificantlyincreasethewaterlevelinlakes.FisheriesandEn-vironmentCanadahasrecordedthatheavyrainfallinJune1973overAlberta caused thewater level ofLakeWinnipeg to riseby1m.AbigriversuchastheSt.Lawrence,withanaverageflowrateof10,100m3/s,wouldtakealmostonemonthtodrainthisadditionalwater(CNC/HID1978;Lawfordetal.1995).TherainfallofearlyJune2002broughtmajorfloodingtotheCana-dianprairies(Szetoetal.2011).FourheavyraineventsinJune2005resultedinfloodinginallprovincesandcausedpropertydamageofapproximately$400million(Shook2016).TheAl-bertafloodofJune2013,causedbyabovenormalspringsnow-melt in theCanadianRockies and extreme rainfall, is one ofthecostliestnaturaldisastersinCanadianhistory,withprojectedcostsofupto$6billion(Pomeroyetal.2016).HencefloodingisconsideredthemostcommonandcostlynaturaldisasterforCanadians.Identificationoffloodproneareasandrestrictionofdevelopmentwithinfloodzonescansignificantlyreducedam-agecausedbyfloodingandcanalsohelptosavelives.
There is a need for updated accurate flood zonemaps ofMoose Jaw to fulfill the requirements of Saskatchewan’sDe-velopmentActof2007.Thisstudydemonstratestheuseofthelatest remote sensing technologies (i.e., high resolution aerialphotographs and point cloudLiDAR) for accurate flood zonemapping,itscomparisonwithtraditionalfloodzonedelimitationgenerated from 500-year historical flood recurrence data, andtemporalanalysisoffloodzonestoemphasizetheneedforregu-larupdatesoffloodzonemapswithinanurbanenvironment.
Figure 1 The study area Cartography: Muhammad Almas
SouthernSaskatchewan isanextensionof theGreatPlainsofcentralNorthAmerica.Muchof the region consists of gentlyrolling hills separated by river valleyswith trees largely con-finedtothesevalleys(Buttleetal.2016).Theregionisprimar-ilyagrasslandheavilyalteredbyagriculturalactivities,andissparselypopulated.MooseJawisamedium-sizedcity,foundedin1882andsituatedabout80kmwestofReginaattheintersec-tionofHighway1andHighway2(Figure1).Itisanimportantrailway junction that connectswithChicagovia theSooLineandisahometo33,890people(StatisticsCanada2016).Down-townMooseJawislocatedincloseproximitytotheconfluenceofThunderCreekandtheMooseJawRiver,andhasahighriskofflooding.Springsnowmeltorheavyrainoverashortperiodof time in the catchment area of SpringCreek also increaseswater levelsandcancauseflooding in somepartsof thecity.
ThecityexperienceditsworstfloodinginApril1974whentheMooseJawRiver,ThunderCreek,andSpringCreekallover-flowedtheirbanks.About60cityblockswereflooded,forcing1400residents toevacuate (MinistryofEnvironmentandCli-mateChangeCanada 2013). Infrastructurewithin the floodedareawasseverelydamagedandessentialservicesweredisrupt-ed.ThisstudyfocusesonfloodzonemappingofMooseJaw.
Traditionally,floodzonemappinginSaskatchewanhasbeenbasedon500-yearhistoricalfloodrecurrencedata, referred toas1:500mappinghaving0.2%(1 in500)chanceofflooding,with additional freeboard for hydrologic andhydraulic uncer-tainties(GovernmentofCanada2013).Hydrologicmodelingis
Alternatively,floodzonescanbemappedthroughidentifica-tionofbasefloodelevation(BFE)whichisdefinedasarefer-encesurfaceelevationbeyondwhichwaterlevelsmaybecon-sideredasfloodedwater.Itcanalsobedefinedastheelevationmarking the edges/boundaries of the floodway. IdentificationofanaccurateBFEisacrucialandcriticalstep infloodzonemappingusing elevation data.Different regulatory authoritiesmayprovideguidelinesforestablishingBFE.Generally,itisacomputedwatersurfaceelevationthatfloodedwaterhasa1%chancetoreachorexceedduringaregularbasefloodevent(Na-tionalResearchCouncil2007).However,dependingupon thetopographyofanarea,itcansimplybeanelevationoftheriveredges,floodwayboundaries,ortheextentofbasefloodevents.OnceBFEhasbeenestablished,anareawithapredefinedeleva-tionaddedtoBFEcanbeclassifiedasafloodplain(Merwadeetal.2008).Dependingupon thegeneral topographyofanarea,localgovernmentsmayrequiredifferentelevationstobeaddedtotheBFEforfloodzonemapping.Asperfloodzonemappingguidelines provided by the Government of Saskatchewan, anareawithasurfaceelevationof0.3mabovetheBFEisconsid-eredfloodplain(Sandinketal.2010).
Theaccuracyofthefloodinundationmapsbasedoneleva-tion information depends upon accurate identification ofBFEand theaccuracyof thedigital elevationmodel (DEM) (Mer-wadeetal.2008).Surfaceelevationforfloodzonemappingistypicallyderivedfromhistoricalcontourmapswhichmaynotprovideaccurateelevationinformationforseveralreasons,no-tablythelimitedspatialandverticalaccuraciesofthesemaps.Highly accurate flood zonemapswith improved spatial reso-lutioncanbe createdusing aDEM.DEMsbasedonNASA’sShuttle Radar Topography Mission (SRTM3), with a spatialresolutionof3arcseconds(≈90mspatialresolution),arefreelyavailable from the United States Geological Survey (USGS)andhavebeenusedbymanyagenciesforfloodzonemapping(vandeSandeetal.2012).NASAhasrecentlystartedofferingDEMs(≈30mspatialresolution)generatedfromSRTMGlobal1arc second (SRTMGL1)data (NASA2017).Thesignificantimprovementof spatial resolution in theSTRMGL1datawillallowuserstoupdateolderfloodzonemaps.Ifavailable,pointcloudLiDARdatacanbeusedtoderivemoreaccurateDEMs,resultinginbetterqualityfloodzonemaps.
Floodzonemapsgeneratedusinghistoricaldataneedtobeupdatednotonlytoimproveaccuraciesbutalsotoincludeup-dated information as thesemaps can become obsolete due totopographic and land-use changes within the floodplain. TheFederalEmergencyManagementAgency(FEMA)oftheUSAestimates thatasofMarch2004,nearly70%ofUnitedStatesfloodmapsweremorethan10yearsoldandwerebasedonout-dateddata(GAO2004).ThesituationinCanadaisnotverydif-ferent.Mostof thefloodplainmaps inCanadawereproducedbetween1976and1997under the federalFloodDamageRe-ductionProgram(FDRP). InMarch2017, the federalgovern-
mentannounceditsintentiontorestarttheprogram(InsuranceBusiness 2017; Natural Resources Canada and Public SafetyCanada 2017).The guidelines provided under the new initia-tiveencourage theuseofpointcloudLiDARdata forextrac-tionofelevationinformationandfloodzonemapping(NaturalResourcesCanadaandPublicSafetyCanada2017).LiDARisoneofseveralmethodsusedtoconstructDEMsandtoderiveel-evationdatafromthem.SeveralstudieshavebeenpublishedtosupporttheuseofLiDARdataforfloodzonemapping(Webster2010).However,nostandardmethodologyhasbeenreportedintheliteraturetodetermineBFEasitishighlyterraindependent.
Methodology
Elevation information was extracted from high density pointcloudLiDARdatacollectedin2014andprovidedbytheCityofMooseJawtogenerateaDEMofthearea.AtypicalairborneLiDARsystemcollecting30points/m2iscapableof5to9cmlateraland5 to19cmverticalaccuraciesdependingupon theflight height. In addition to the flight height, accuracies of aLiDARgeneratedDEMalsodependuponthetopographyandlandcoverof thearea (HodgsonandBresnahan2004).Accu-racyshouldbecheckedbyusingasufficientnumberofgroundcontrolpoints(GCPs).TheAmericanSocietyforPhotogramme-tryandRemoteSensing(ASPRS)andtheInter-GovernmentalCommitteeonSurveyingandMapping(ACSM)haveproposedtheuseofaminimum30GCPsforaccurateassessmentofLi-DAR-derivedelevationdata(Pouralietal.2014).
ThecombineduseofhighresolutionaerialphotographsandpointcloudLiDARdataisemployedinthisstudytoidentifyriv-eredges,floodwayboundaries,andthesignaturesofbasefloodevents.Acombinationofthesefeaturesisthenusedtodeterminea river profilewhen establishingBFE.Theproposedmethod-ologyaddsasafemargin to theestablishedBFEtodeterminea floodplain. This objective was achieved through extractionof elevation contours at regular interval fromLiDAR-derivedDEM.ArcMAP10.4offersabuiltintoolunderSpatialAnalystand3DAnalystextensionstogenerateelevationcontoursfroma raster DEM image.This tool was used to extract elevationcontours.Differentverticalintervalsforelevationcontoursweretried,however,contoursof0.25verticalintervalwerefoundthemostsuitableastheycouldberelatedeasilywithBFE.Theothercontourintervalsproducedeither toodenseor toosparsecon-tourlinesandweredeemedunsuitable.TheverticalaccuracyofthesecontoursdependsupontheaccuracyoftheLiDAR-derivedDEMandwasverifiedusing100welldefinedGCPscollectedusingasurveygradeTrimble6000seriesGeoXHGPSdevicehavingsub-cm3Daccuracy(Raoufetal.2017).
The rootmean square (RMS)errorof theLiDAR-derivedelevation contours is usually used to calculate for accuracyassessment of elevation data (Gianinetto andFassi 2008).AnRMSerror of 0.15mwas observed in the elevation contoursof theLiDAR-derivedDEM.As indicatedearlier,accuracyofLiDARelevationdataishighlyterrainandlandcoverdependent(HodgsonandBresnahan2004).Consideringthefrequenteleva-
Highresolutionaerialimagerytakenin2014withaspatialresolution of 14 cm,was used formappingwater bodies andotherlanduseinMooseJawandthecity’simmediateenvirons.Thethreewaterchannelswithinthecitylimitshavedistinctivesharply carved edgeswith steep slopes.Only a small portionoftheMooseJawRiverhasawiderchannelwhereriveredgesarenotsharplycarved.However,regularbasefloodeventshascaused erosion and left permanent signatures, thusmaking iteasy to identify in thehighresolutionaerial images.Differentimageclassificationtechniquescouldhavebeenappliedtoex-tractthisinformationbutresearchsuggeststhattheuseofhands-ondigitizationisapreferredtechniqueforsuchfeaturesusinghigh resolution aerial photographs (Brown and Young 2006;Raouf et al. 2017).Although time consuming, hands-on digi-tizationprovidessuperiorandaccurateresultsforriverprofileidentification.Elevationcontoursof0.25verticalintervalweresuperimposedontheclassifiedaerialimagerytofindelevationinformationforthethreewaterchannelprofiles.ThiselevationinformationwasusedtodeterminetheBFE.
Three dams are constructed on the Moose Jaw River toregulatethewaterdischargeandtocontrolfloodingwithintheproject area.Theelevationof the riverprofile is considerablydifferent insectionsbeforeandaftereachdam.Theriverwasdividedintothreesectionstoaccommodatethesechanges,andanaverageelevationvaluewasusedtodeterminetheBFEforeachsection.Elevationcontourswithanelevation0.5mhigherthanBFEwereusedforfloodzonemappingofeachsection.Inadditiontoidentifyingtheriverprofile,otherinfrastructuresuchasroadsandbridgeswereidentifiedintheaerialimages.Theel-evationsofthesestructureswererecordedtodeterminetheiref-fectonfloodzonedelimitation.Recordingthesedatarecognized
thatelevatedinfrastructurecanactasaprotectiveboundaryforfloodsandcanchange thefloodzonepattern in thearea.TheoverallmethodologyoftheresearchissummarizedinFigure3.
Itwas observed that the elevation contours of 0.25m in-terval, superimposedon the riverprofiles,werenot100%co-incidentwith themanda fewgapsbetween the twowereob-served.Thesegapswereusedtoselectobservationpointsandthe elevation of the river profile at these pointswas recordedusingahandheldsurveygradeTrimble6000seriesGeoXHGPSdevice.Nosignificantchangewasobservedalongdownstreamriver profile elevations of Thunder Creek and Spring Creek.Anaverageofallelevationchangewasusedtodetermine theBFEoftherespectivecreeks.However,riverprofileelevation
Figure 4 A comparison of flood zones generated from LiDAR data and 500-year flood recurrence data Cartography: Muhammad Almas
changedmarkedlybeforeandafterthethreedamsontheMooseJawRiver.ThreedifferentBFEvalueswereusedtoaccommo-date theseelevationchanges in the riverprofile.Theuseof adifferentBFEvalueforeachsectionallowedaccuratefloodplainmappingalongdifferentsectionsoftheMooseJawRiver.Whilethesedamscanbeusedforfloodprevention,theycanalsocauseflashfloodingdownstreamifthegatesareopenedsuddenlytoprevent the dam from collapsing because of excessivewater.Theinletsofthedamsalongwiththebridgesarepotentialareasforicejamoccurrencesalongtheriver.Icejamsatthesepointsofflowconstrictioncancauselocalfloodinganddamagetothedamsandbridges.Thelocationsofthedamsandbridgeswereextracted from the high resolution aerial photographs. Eleva-tionsofwaterclearancesunderthebridgesanddamelevationswererecordedusingahandheldsurveygradeGPSdevice.TheBFEvalueforeachsectionoftheriverprofilewasadjustedtoaccommodatetheclearanceelevationsof thebridgesanddamelevations.TheextentofthefloodzonegeneratedfromLiDARdatawascomparedwiththetraditionalfloodzonebasedon500-
Figure4showsthatbothzonesweremostlysimilarinex-tent, however, a considerabledifferencebetween the twowasobserved at some locations. The LiDAR-derived flood zonesarelessextensivethanthosebasedon500-yearhistoricalfloodrecurrencedataduetotheirincreasedaccuracyandabilitytoac-commodatechangesinlandcover.ThesedifferencesweremostpronouncedalongThunderCreek.Land-coveranalysisofaerialimageryrevealedthatinfrastructuredevelopmentintheareaisoneofthemaincausesofthesealterationstothefloodplains.Inparticular,Highway2,Highway363,ManitobaExpressway,9thAvenueSWandHighStreetWestwereraisedconsiderablytohigherelevationsduringconstructionandrepaving.Theseroadsactasfloodprotectionbarriersandhavechangedthefloodzoneswithin the city boundaries. However, not all the roads wereraisedtothehigherelevationsandarestillatgreatriskoffloodwaterinundation.Itwascalculatedthat19.8kmofcityroadsliewithinthefloodzones,haveahighriskofflooding,andshould
Technologicaladvancementsindatacollectionandmappingtechniqueshavecontributed to recordedchanges infloodplaingeometry.Traditionalflood zonemaps are based on 500-yearhistorical flood recurrence data.However, data collection andmappingtechniqueshavechangeddrasticallybecauseoftech-nologicaladvancements. Inaddition,cartographicdatacollec-tionscaleandthedigitizationofhistoricalmapscanintroduceerrors.Thesearedifficult todeterminewithprecisionbecauseofmissinginformationregardingdatacollectiontechniquesandtheequipmentusedwhencollectingthedata.Inherentinaccu-raciesofhistoricaldataandinfrastructuredevelopmentwithinflood zones are the main causes of changes within the floodzones.Acomparisonofthetwofloodzonesalongthethreewa-terchannelswithintheprojectareaissummarizedinTable1.
Use of the latest remote sensing technologies such as Li-DARforfloodzonemappingoffersadditionalbenefits.TheLi-DARdatacombinedwiththehighresolutionaerialphotographscan be used to calculate building footprints and impermeable
surfaceareawithinthecitylimits.Sincewatercannotpenetratethroughthesesurfaces,almost100%of the incidentrainwaterrunsoffintothecitydrainagesystem.Thisinformationcanbeusedforhydrologicmodelingandthedesigningofflooddrain-agesystemsfor thecity.Abuildingfootprint imagegeneratedfromtheLiDARdataisalsoshowninFigure5.
The calculated area of the building footprints within citylimits was 289.9 ha. Similarly, paved roads and parking lotswereextractedfromthehighresolutionaerialphotographs.Thecalculatedareaofpavedparkinglotswas144.5haandthetotal
lengthofpavedroadswas284.5km.Thisareacanhaveasig-nificant contribution to the accumulationofwater during rainstorms and requires adequate drainage systems to protect thecity from inundations during intense rainfalls and consequentflashflooding.
Conclusions
ThestudyconcludedthatLiDARdatacanbeusedsuccessfullyto generateDEMs in an urban environment and to derive el-evationcontoursfromtheDEMusinginterpolationtechniques.However, the accuracy of these elevation contours should beverified using well distributed GCPs before using them forflood zonemapping. It is also concluded that the informationobtainedonlyfromLiDARgeneratedelevationcontoursisnotsufficientforfloodzonemappingandmustbesupplementedbyhigh resolutionaerialphotographs to identifyflood signaturesandwaterbodieswhichcanprovideabasisfordeterminingBFE.Thesehighresolutionaerialphotographscanhelpintheidenti-ficationoffloodbarriersincludingdamsandbridges.TheBFEshouldbe adjusted to accommodate the elevation informationofthesefloodprotectionstructuresandpotentialfloodbarriers.Thisinformationismostusefulinrefiningfloodzonesderivedfromelevationdata.Floodzonesgenerated fromLiDARdatacanbesignificantlydifferentfromthefloodzonesbasedonhis-torical flood data, especially in an urban environment, as de-velopmentofinfrastructurecanalterfloodzones.Thisrequiresregular updating of flood zonemaps andLiDARdata can beusedsuccessfullyforthispurpose.Inadditiontothefloodzonemapping,LiDARdata can be used formapping impermeablesurfaceswithinanurbanenvironment.Thisinformationcanbeusedforhydrologic/hydraulicmodelingtodesigneffectiveandadequate flood drainage systems. The study has successfullydemonstratedan integrateduseofvarious remotesensingandGIStechnologiestoupgradethefloodzonemapsofMooseJaw.Similartechniquescanalsobeusedtoupgradethefloodzonemapsofothercities.
Buttle,J.M.,D.M.Allen,D.Caissie,B.Davison,M.Hayashi,D.L.Peters,J.W.Pomeroy,S.Simonovic,A.St-Hilaire,andP.H.Whit-field.2016.Floodprocesses inCanada:Regionalandspecialas-pects.Canadian Water Resources Journal41(1–2):7–30.
CNC/HID(CanadianNationalCommitteefortheInternationalHydro-logicalDecade).1978.Hydrological atlas of Canada.Ottawa,ON:FisheriesandEnvironmentCanada.
GAO.2004.Flood map modernization: Program strategy shows prom-ise, but challenges remain.Washington,DC:UnitedStatesGeneralAccountability Office. https://www.gao.gov/assets/ 250/241972.pdf.
Gianinetto,M.,andF.Fassi.2008.ValidationofCartosat-1DTMgen-erationfor theSalondeProvence testsite.The International Ar-chives of the Photogrammetry, Remote Sensing and Spatial Infor-mation Sciences37(B1):1369–1374.
GovernmentofSaskatchewan.2007.The Planning and Development Act, 2007. Regina, SK:TheQueen’s Printer. http://www.qp.gov.sk.ca/documents/english/statutes/statutes/p13-2.pdf.
Hodgson,M.E.,andP.Bresnahan.2004.AccuracyofairborneLiDAR-derivedelevation:Empiricalassessmentanderrorbudget.Photo-grammetric Engineering and Remote Sensing70(3):331–339.
Insurance Business. 2017. Updated flood maps coming to Canada.http://www.insurancebusinessmag.com/ca/opinion/updated-flood-maps-coming-to-canada-78259.aspx.
Lamond,J.E.,R.D.Joseph,andD.G.Proverbs.2015.Anexplorationoffactorsaffectingthelongtermpsychological impactanddete-rioration ofmental health in flooded households.Environmental Research140:325–334.
Levin, J.N.,A-M.Esnard, andA.Sapat. 2007.Populationdisplace-mentandhousingdilemmasduetocatastrophicdisasters.Journal of Planning Literature22(1):3–15.
Merwade,V.,F.Olivera,M.Arabi,andS.Edleman.2008.Uncertaintyinfloodinundationmapping:Currentissuesandfuturedirections.Journal of Hydrologic Engineering13(7):608–620.
NaturalResourcesCanada andPublicSafetyCanada. 2017.Federal floodplain mapping framework. Ottawa, ON: Natural ResourcesCanada.https://doi.org/10.4095/299806.
Pomeroy, J.W.,R.E.Stewart, andP.H.Whitfield. 2016.The2013flood events in the South Saskatchewan and Elk River basins:Causes,assessmentanddamages.Canadian Water Resources Jour-nal41(1–2):105–117.
Pourali,S.,C.Arrowsmith,N.Chrisman,andA.Matkan.2014.Verti-calaccuracyassessmentofLiDARgroundpointsusingminimumdistanceapproach.InProceedings of Research at Locate’14,ed.S.Winter,andC.Rizos.Canberra,Australia:Research@Locate’14:86–96.
Thistlethwaite,J.,D.Henstra,S.Peddle,andD.Scott.2017.Canadian voices on changing flood risk: Findings from a national survey.Waterloo, ON: Faculty of Environment, University ofWaterloo.https://uwaterloo.ca/climate-centre/sites/ca.climate-centre/files/up-loads/files/canadian_voices_on_changing_flood_risk_fnl.pdf.
