Table1.DetailedDescriptionofthe14AdvancedSpaceborneThermalEmissionandReflectionRadiometerInstrumentBandswithTheirWavelengthsandResolutionsInstrumentVNIR*SWIR*TIR*Bands1–34–910–14Resolution

15m30m90mCross-trackpointing±318km±116km±116km(±24°)(±8.55°)(±8.55°)Swathwidth

60km60km60kmQuantization(bits)

8

8

12

*VNIR=visiblenearinfrared;SWIR=short-waveinfrared;TIR=thermalinfrared.

thePlaneTracetoolineithera2-ora3-Dview,sedimentarybeddingorfaultplanesareidentifiedandoverlainbytransparentvirtualplanes.Thevir-tualplanesaretheninteractivelyadjustedtotheorientationsthatbestrepresentthetracesoftheintersectionofthegeologicfeatureswiththeDEM.Evenwithoutexposedrocks,thedipdirectionanddipanglecanbemeasuredapproximatelyifitcanbereasonablyassumedthatthetopographyreflectsthedistributionoftheunderlyingsedimentarybed-ding.Inthenextstep,thesizeofthevirtualplaneisadjustedtorepresentasectionofthegeologicstructure,whichcanbeapproximatedbyaplane(Figure5B).Finally,theparametersofthebestfitvirtualplanecanbesavedforfurtherprocessinginthreedimensions(e.g.,interpolationofnon-planarsurfaces;Figure5C)orthedipanglescanbeprojectedintoa2-Dprofiletoconstructbal-ancedcrosssections(Figure5D).Asthedataareorganizedinseparatelayers(i.e.,geologicmap,falsecolorsatelliteimage,dipmarks,faults,seismiccrosssections),theonesthatshouldbeprojectedontothecrosssectioncanbechosen.

STRUCTURALMODELINGOFTHEZAGROSFOLDANDTHRUSTBELT,NORTHEASTERBILTotestthePlaneTracemodule,wecalculateddipdirectionsandanglesatmorethan100lithologicboundariesinthefoldtrainsnortheastofErbil.Theaimofthiswastwofold.First,wecomparedthecalculatedspatialorientationswithmeasure-mentsofthesamelithologicboundariesinthe

fieldtotesttheaccuracyofthetechnique.Sec-ond,thedatawereusedtoconstructastructuralmodel,includingasimplebalancedcrosssection,ofanareanotinvestigatedindetailinthefieldbecauseofsecurityconcerns;thiswastodemon-stratetheapplicabilityofthemethod.

Alldata(advancedspacebornethermalemissionandreflectionradiometer[ASTER]andSPOT[satellitepourl′observationdelaterre]satelliteimagery,geologicmap)wereloadedintoWinGeol

Table2.CorrelatedDipDirectionsandDipAnglesofFieldMeasurementsandTheirDigitalElevationModel–DerivedEquivalents*Location

3a

3b

3c

3d4a4b

4c

Deriveddipdirection2242232262262454136.97Measureddip2242182242242253634direction

Deriveddipangle65704040476560Measureddipangle46784040456459Location

4d4e4f

5a

5b

5c

Deriveddipdirection343742.12203225233Measureddip353630180207225direction

Deriveddipangle493014282440Measureddipangle504024292544Location

6a

6b

7a

7b8a

9a

Deriveddipdirection34318.523223278221Measureddip3136023222474203direction

Deriveddipangle403564474077Measureddipangle33466450

4079

Location

9b

9c10a11a11b11cDeriveddipdirection22128170230.1250250Measureddip20730170213225240direction

Deriveddipangle792220251111Measureddipangle79

16

24

37

1815

*Thelocationnameconsistsoftheareanumber(1–12)accordingtoFigure3andtheletterrepresentingtheindividualmeasurementlocationstartingfromthenorthwesttothesoutheast(seeFigure6).

Reifetal.949

Figure6.(A)ASTER(advancedspacebornethermalemissionandreflectionradiometer)digitalelevationmodelwiththedifferentstratigraphicformationsoftheinvestigatedareashowingthelocationofthe25testsites,wherefieldmeasurementsandcalculatedorientationshavebeencompared.(B)ScatterplotcomparingthefieldmeasurementsandmeasurementsacquiredusingPlaneTracefrom25locationsinthestudyarea.Theyaxisshowsthedihedralangle(givenviathedotproductofthenormalsoftheseplanes)betweenthemeasuredplanarstructureandcalculatedplane;thexaxisshowstherealdipangleofthemeasuredplanarstructure.Theredandbluepointssymbolizemeasurementsfromanticlinesandsynclines,respectively.

andcombinedwiththeDEMtocalculatethespa-tialdirectionsoflithologicboundariesusingPlane-Trace.Severalspectralbandcombinationsofdif-ferentsensors(V1,V2,andtheinfraredV3nandV3b)providedbytheASTERsatellite(Table1)

Table3.EvaluationofCorrelationbetweenField

MeasurementsandDigitalElevationModel–DerivedValuesat25LocalitiesthatShowstheTimeandAccuracyVariationoftheMethod*Value

AverageerrorStandarddeviationAmountofexactmeasurementsAveragetime

Maximaldifference

DimensionsDecimaldegrees

–nof25Minutes

Decimaldegrees

DipDirection9.7010.2637.6641.48

DipAngle4.564.895–19

*Theaveragetimerequiredisstatedforonecompleteprocess(i.e.,onedipdirectionanddipanglemeasurement).

wereusedtoenhancethecontrastofthestratigraphicformations,focusingonboundariesbetweencom-petentandincompetentlithologies.Theexcellentoutcropsandthelowvegetationcoverfacilitatedanaccuratevisualizationofthevirtualplaneontothelithologicboundaries.

Thestructuralfieldmeasurements(measuredwithageologicfieldcompassandbasedonatleast10independentmeasurementsatthatsite)havebeencomparedwithspatialorientationsderivedbythePlaneTracetool(Table2)at25locations(Figure6).Comparisonofthesevaluesrevealsthatthedipdirectionsmeasuredinthefieldwerere-producedbyPlaneTracemeasurementswithinanerrorof±9°,andthedipanglesderivedfromtheremotesensingdatadeviatefromthefieldmea-surementswithinanerrorof±5°(Table3).

Thepresentedmethodhasbeenusedtode-riveapproximately100measurementsfromre-motesensingdata,sufficienttoconstructasimplebalancedcrosssectionthroughthearea.Becausetheprofilewasnotinvestigatedduringfieldwork,

950QuantitativeStructuralAnalysis,Kurdistan,Iraq

Figure7.(A)ASTER(advancedspacebornethermalemissionandreflectionradiometer)digitalelevationmodel(DEM)withthedifferentstratigraphicformationsoftheinvestigatedareashowingthelocationofthecrosssectionCC′andthedipdirectionanddipanglemeasurementscalculatedwithPlaneTrace(themeasurementsusedfortheconstructionofthecrosssectionCC′areshownasredinclinedplanescuttingtheDEM).ThefaultmeasuredusingPlaneTraceandtheprojectionofthebalancedcrosssection(dottedline)areshowninblueandyellow,respectively.(B)CrosssectionCC′constructedonlyfromthecalculateddipangleswithPlaneTraceusingtheconstantdipdomainmethod(TearpockandBischke,2003).

Reifetal.951alltheavailableorientationdataarederivedfromtheDEM,enhancedbydrapedsatelliteimagesandthegeologicmap(BudayandJassim,1984).Anin-clinedviewofthecrosssection(Figure7A)showsthemeasurementlocations(indicatedasredplanes)thathavebeenprojectedontothecrosssection(Figure7B).Thegreatestprojectiondistanceofthemeasurementsfromthecrosssectionwassetintheprograminterfaceto5000m(16,404ft).ThecorrectionofapparentdipanglesformeasurementswithadipdirectionobliquetothecrosssectionisautomaticallyperformedbyPlaneTrace.Withtheadditionalinformationofthegeoreferencedgeo-logicmap,theconstantdipdomainmethodwasusedtoconstructasimplebalancedcrosssection(TearpockandBischke,2003).Thisidentifiessec-torsofsimilardipanglesacrossafoldedregionandseparatesthesesectorswithaxialplanesbisectingthedipanglebetweenneighboringdomains.Byin-cludingthemappedlithologicboundaries,asim-pleline-andarea-balancedcrosssectionhasbeenconstructed.Aspreviouslynoted,theinvestigatedareaisdominatedbyopenuprightfolding,withinwhichnomajorfaultshavebeenidentified.NoevidenceintheremotesensingdatawasseenforthethrustfaultmarkedonthegeologicmapbyBudayandJassim(1984).However,althoughitwasnotrequiredtodrawabalancedcrosssection,ithasbeenincludedasalow-anglestructurewithaninferredminordisplacement.

DISCUSSION

Traditionally,mappingtechniqueshaveusedstruc-turecontourstoconstructintersectionsofgeologicstructureswiththetopographyonmaps.However,evenifthegeologicstructureisapproximatedbyaplaneandthestructurecontoursarethereforeequallyspacedparallellines,thistechniqueistimeconsumingandinaccuratebecausetheerrorsarecumulative.

The30-m(98-ft)resolutionoftheworldwideavailableASTERdatasetsusedwassufficientforestimatingthedipdirectionanddipangleofthelargerscalestructures.Atlocalitieswhereonlystruc-tureslessthan30m(98ft)insizeorveryindistinct

952

QuantitativeStructuralAnalysis,Kurdistan,Iraq

structureswerepresent,itwasnecessarytouseSPOTimagestorevisetheestimations.Withoutthese,estimatingthesurfaceorientationwasnotpossible.Theuseofmoredatawithhigherreso-lution,suchasSPOT(resolution1.0–2.5m[3.3–8.2ft])orQuickbird(high-resolutioncommercialearthobservationsatellitelaunchedin2001withtheresolutionasmuchas0.6m[2ft]),wouldmakethemethodevenmoreusefulinthemajorpartoftheareaofinterest.Comparedwithotherstudies(Snideroetal.,2009),PlaneTracegavegoodresults,despiteusingmostlylowerresolutiondatathanusualforthistypeofwork.Useofhigherresolu-tionDEMandsatelliteimagerylikeSPOT-HRG(high-resolutiongeometricinstrumentswiththeresolutionasmuchas2.5m[8.2ft])canevenim-provetheresults.However,theaccuracyofthemethodisnotonlydependentontheresolutionofsatelliteimagesoraerialphotography,butcanbealsostronglyinfluencedandlimitedbyotherfac-tors,suchasgeomorphology,vegetationcover,andthecomplexityofthetectonicdeformationhistory.Incaseofdensevegetation,themethodwillstillworkaslongassignificantvariationsinelevationcanbeobserved.Lithologicunitsmightbecov-ered,butinsome(rare)cases,evenamplifiedbysupportingthegrowthofdifferentplanttypesorvariablesupportofplantgrowth.Soil-coveredand/ordeeplyweatheredzonesespeciallyinflatareasareofcoursewhereweexpectthebiggestdifficul-tieswiththismethod.Acarefulimageprocessingmighthelpincreasethecontrastofhiddengeologicunits,andtohelpinterpretation,itwouldbeusefultostartfromareaswithlessunfavorableconditionsandtoworktowardtheproblematiczones.Theaccuracyofthemethoddependsofcourseontheexposedelevationdifferences.

ThemeasurementswereshowingthatPlane-Traceismoreaccurateinthesynclinesthanintheanticlines.Thisdiscrepancywascausedbyahighernumberofpoor-qualityoutcropsintheanticlines(threeveryinsignificantoutcropswithanangledifferencehigherthan10°)thaninthesynclinesandnotbythemeasuringalgorithm.Incaseofthesamenumberoftheoutcropswiththesamequalityintheanticlinesaswellasinthesynclines,therewillbenodifferenceintheaccuracy.