For the past 12 years, Cornell SyriaProject scientists and colleagues at theSyrian Petroleum Company havestudied the regional structure and geo-logic evolution of Syria. We are cur-rently generating new structural mapsand tectonic models for the wholecountry. Information on this region isrelatively limited, despite the localimportance of hydrocarbon produc-tion and abundant surface and sub-surface data. Our regional approachinvolves new interpretations of seismicreflection profiles, well data, remotesensing imagery, and potential-fielddata, merged with existing interpreta-tions of similar data sets. These inter-pretations, integrations, analyses, andmap preparation are all performedwithin a GIS platform.

As detailed elsewhere in this issue,the importance of GIS as a data stor-age and interrogation tool for petro-leum exploration is well established.This article describes our use of GIS tofacilitate regional tectonic mapping inSyria. Although not directly related tothe search for hydrocarbons, the mapsand models generated have obviousutility for oil exploration. Herein wedetail the types of data being used,their integration and interpretationwithin the GIS, and our preliminaryanalysis and findings. We will showhow a GIS approach eases data archiv-ing and map generation and also pro-vides interpretational possibilities notavailable with more traditional map-ping procedures.

Our interpretations show thatmuch tectonic deformation withinSyria occurred within repeatedly reac-tivated “mobile zones.” GIS-generatedstructural maps on various geologichorizons, and tectonic maps at keytime points, illustrate the intercon-nected and contemporaneous natureof the tectonic episodes within the dif-ferent Syrian mobile zones.Hydrocarbon plays in Syria, such asthe Euphrates Fault System, owe theirexistence to these tectonic episodes;better understanding of the deforma-tion will lead to more informed explo-ration strategies. Hence, the datastored within the GIS can directly

influence exploration strategies anddecisions.

Study area and GIS coverages. Ourresearch concerns the northernArabian Platform, specifically Syriaand immediately surrounding areas(Figure 1). Past research by CornellSyria Project scientists concerns thetectonic evolution of the different tec-tonic zones within Syria, as delineatedin Figure 1. The previous studies estab-lished the timing and style of structuraldeformation within each zone.Previous attempts were made to cor-relate the findings; however, these pastsyntheses relied on fewer data thancurrently available, and a full integra-tion of all available data and resultswas never attempted. In the currentwork we are taking a truly regional

approach, focusing simultaneously onall parts of the country, notwithstand-ing nonuniform data distribution.

To achieve a fully regional andcomprehensive tectonic analysis ofSyria, we are constructing data “cov-erages” or “layers” within a GIS. Thegeographic extent of each coverage isat least all of Syria. There are threetypes of coverage in the GIS: Thosethat contain interpretable geosciencedata, those that do not, and those thatare interpretations of data. The cover-ages that do not contain geosciencedata are mainly of the “data location”type. For instance the “SeismicReflection Profiles” coverage does notcontain all the seismic reflection databut merely the location and parame-ters of those data (much of the seismicdata base is available only as paper

176 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 00

Tectonic map and geologic evolution of Syria: The role of GIS

GRAHAM BREW and MUAWIA BARAZANGI, Cornell University, Ithaca, New York, U.S.TARIF SAWAF, Syrian Petroleum Company, Damascus, SyriaKHALED AL-MALEH, Damascus University, Damascus, Syria

Figure 1. Topographic image of the northern Arabian Platform. Dashed redlines outline active tectonic “mobile zones” that are named in red. “Stable”platform blocks are in white. Also shown is the distribution of seismicitywithin the northern Arabian Platform as taken from the InternationalSeismological Center (ISC) database for 1964-1994 (see Table 1). Note thatthe apparent lack of seismic activity along the Dead Sea fault system inwestern Syria is likely a consequence of uneven seismic station coverage.

records, and scanning these recordsinto the GIS remains a task for thefuture). Table 1 details the coverageswithin the GIS we are creating, dis-tinguishing among the three types.Some metadata (data about the data)for the coverages are also shown inTable 1, including original source andscale, quantity or extent of the cover-age (e.g., kilometers of seismic data),and so on. We are using ESRI(Environmental Systems ResearchInstitute) products, namely Arc/INFOand ArcView, for most of our work, sothis discussion follows ESRI nomen-clature.

Understandably, the coverageshave varied provenance. Some havebeen digitized at Cornell from pub-lished sources, some are downloadedfrom Internet archives, and others arethe products of our interpretations.During the construction of the cover-ages, strict quality control ensures thehighest accuracy both within andamong data sets. This means estab-lishing original map projection parameters, careful digitization, metic-ulous registration, and cross-checks.

Presently we use the Levant Lambertprojection for most of our work,although a reprojection of the cover-ages is trivial.

Storing data within a GIS hastremendous advantages in compari-son with hard-copy records or inde-pendent computer data bases. Onceall the data are entered, data accessbecomes much faster. For instance,finding all gamma-ray logs recordedwithin Silurian shale or locating allseismic lines processed beyond 4 s

two-way time becomes elementary.Similarly, adding new data is quickand easy, with the impact of new infor-mation on any given interpretationinstantly accessible. Fence diagrams,isopachs, grids, cross-sections, anima-tions, and other interpretable productsare easily created. Furthermore, asshown in the following examples, aworkstation-based GIS interpretationpermits a 3-D appreciation of the sit-uation and frees the user from the con-straints of a traditional “profile-based”

0000 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 177

Figure 2. Composite image of selected coverages available at Cornell from within the GIS including a 3-D perspec-tive view of Syria showing data locations.

Seismic Industry Basemaps 1:200 000 ~17 000 km Arcs and Points –ReflectionSeismic Basemap 1:500 000 ~875 km Arcs –

Refraction

COVERAGE COVERAGE SOURCE/ ORIGINAL QUANTITY/ FEATURE SPECIALTYPE REFERENCE SCALE EXTENT TYPE NOTES

Table 1. All coverages currently included, or planned, for our GIS platform at Cornell. As the work progresses,some details of this list are likely to change. For the stratigraphy coverages, our initial intent is one coverage foreach Mesozoic formation. DCW = Digital chart of the world.

SeismicData

Locations

International DCW 1:1 000 000 Global Arcs –Borders www.esri.com

Urban Areas DCW 1:1 000 000 Global Polygons –www.esri.com

Roads DCW 1:1 000 000 Global Arcs –www.esri.com

Rail Roads DCW 1:1 000 000 Global Arcs –www.esri.com

Lakes DCW 1:1 000 000 Global Polygons –www.esri.com

Rivers DCW 1:1 000 000 Global Arcs –www.esri.com

CulturalData

Field Various 1:500 000 Syria Polygons Production,Locations Pay Zone etc.Pipelines Various 1:500 000 Syria Arcs –

Hydrocarb.Production

Well Locations Various Various Nearly 400 Points Selected Log Dataand Formation Tops Wells to be included

Facies Well Logs – Selected Wells – From WellInformation from Syria Logs

Well Data

Bouguer Gravity BEICIP (1975) Various Syria, Lebanon Grid and Arcs 2 mGalAnomalies Resolution

Aeromagnetic Filatov (1958) 1:500 000 Syria Grid and Arcs 25 nTAnomalies Resolution

PotentialField Data

Landsat Eros Data Center ~30 m Ground See Figure 4 Grid Seven SpectralThematic Maper edcwww.cr.usgs.gov Resolution Bands

Topography GTOPO30 ~1 km Ground Global Grid BathymetryGesch et al. (1999) Resolution also Available

Imagery

Geologic Field Mapping 1:500 000 Syria, Turkey, Polygons –Map Ponikarov (1966) Lebanon

Seismicity ISC Dataset – ~22 000 events Points –www.isc.ac.uk for study Area

Earthquake Focal Harvard CMT – – Points For MagnitudesMechanisms www.seismology.harvard.edu >~5

Volcano Locations Ponikarov (1966) 1:500 000 Syria, Turkey Points Ages alsoand Others Available

Detailed Various 1:1000 Palmyrides/ – Detailed SurfaceField Observations Coastal ranges SectionsTectonic Domain Cornell – Northern Arabian Polygons –

Boundaries Interpretation PlatformSurface Folding Ponikarov (1966) and 1:500 000 Syria Arcs –

Image InterpretationsSurface Faulting Ponikarov (1966) and 1:500 000 Syria Arcs –

Image Interpretations

Geology /Tectonics

Top Paleogene Cornell 1:500 000 Syria Grid and Arcs Fault Activity,Interpretations throw etc.

Top Cretaceous Cornell 1:500 000 Syria Grid and Arcs Fault Activity,Interpretations throw etc.

Top Lower Cornell 1:500 000 Syria Grid and Arcs Fault Activity,Cretaceous Interpretations throw etc.Top Triassic Cornell 1:500 000 Syria Grid and Arcs Fault Activity,

Interpretations throw etc.Top Paleozoic Cornell 1:500 000 Syria Grid and Arcs Fault Activity,

Interpretations throw etc.Top Mid-Cambrian Cornell 1:500 000 Syria Grid and Arcs Prominent

Interpretations Seismic ReflectorTop Metamorphic Cornell 1:500 000 Syria Grid and Arcs Predominantly from

Basement Interpretations Refraction Analysis

StructuralMaps

Isopachs Interpreted 1:500 000 Syrian Mesozoic Arcs FutureSection Project

Facies Interpreted 1:500 000 Syrian Mesozoic Arcs FutureVariation Section Project

Paleogeography Interpreted 1:500 000 Syrian Mesozoic Arcs FutureSection Project

Source Rock Interpreted 1:500 000 Syrian Mesozoic Arcs FuturePotential etc. Section Project

Sed./Strat.Analysis

NO

N-D

ATA

BE

AR

ING

DA

TA B

EA

RIN

GS

INT

ER

PR

ETA

TIO

NS

178 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 00

analysis.Figure 2 is a simple demonstration

of how selected coverages can be visu-alized. On a workstation, these per-spective views are far more revealingthan this paper reproduction; userinteraction allows examination fromany angle, with appropriate process-ing to highlight specific features. Thetop of Cretaceous surface was made bygridding only well data. Because seis-mic and other data have yet to beincorporated into this surface, faultsare not shown. Relationships amongtopography, subsurface structure, andgravity response are apparent even inthis preliminary view.

The utility of 3-D montages suchas Figure 2 goes far beyond simplepresentations. Figure 3 shows how sur-face, subsurface, and potential-fielddata may be used together. This typeof GIS visualization allows observationof along-strike variations and subse-quent modifications that are not pos-sible when working on profile-typemodels. Another use of a 3-D per-spective view might be to show sur-face geology draped over topography,thus illustrating patterns of erosion.

As part of the current regionalwork, we are interpreting tectonic fea-tures visible on Landsat ThematicMapper (TM) images to aid structuralmapping. GIS is particularly usefulwhen interpreting remote sensingimagery, as the image can be viewedon screen and coregistered coveragesoverlain to aid interpretation. Forexample, known faults and foldsmapped in the field can be overlainalong with surface geology to establishthe spectral and spatial characteristicsof the various features and hence guidethe interpretation. High-resolutiondigital topography can further guidefault mapping. Topography can beused with bedding traced fromimagery to calculate regional beddingattitudes (the “three-point problem”).

The Cornell University library ofLandsat Thematic Mapper images forthe Middle East includes the equiva-lent of more than 115 individual TMscenes. Many others are available forNorth Africa and other locations(Figure 4). For further information seehttp://atlas.geo.cornell.edu/htmls/TM/mena.html.

Construction of structural maps.Construction of the structural mapslisted in Table 1 is at a scale of 1:500000. Figure 5 shows a preliminaryexample. The contours of depth toLower Cretaceous are based on welldata, together with digitized, depth-

converted, seismic-reflection interpre-tations that have been entered into theGIS. Faults are interpreted from seis-mic reflection profiles and other data

and their digitized locations enteredinto the GIS. Both the depth contoursand the faults can be edited on screen.In this way, coregistered topography,

0000 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 179

Figure 4. Preview image of some Middle East Landsat TM images stored atCornell GIS. Red lines indicate locations of other images at Cornell.

Figure 3. TM imagery draped on topography of the Ghab Basin region (seeFigure 1), northwest Syria, and basin morphology. The lower surface hasbeen lowered relative to its true position for ease of viewing. The Bouguergravity contours generally exhibit distinct lows near the deepest portionsof the basin. The occasional offset between gravity lows and basement isowing to depocenter migration with time. Note the asymmetric depositionwithin the basin and the segmentation of the deposition into smaller sub-basins.

180 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 00

Figure 6. Preliminary maps showing schematic tectonic evolution of Syria. The six time frames illustrate dominanttectonic events and features at each of those time points. Once complete, the GIS will hold sufficient attributes toautomatically generate maps such as these.

Figure 5.Preliminarystructuralmap show-ing the topof LowerCretaceoushorizon.This is oneof a series ofsubsurfacestructuralmaps underconstructionwithin theGIS (seeTable 1).

satellite imagery, and geology mapscan facilitate mapping of features.

Each fault, such as those shown inFigure 5, will be assigned extensiveattributes within the coverage. Amongthese will be the probable timing andsense of motion on each fault. Thisinformation is important becausemany faults within Syria have experi-enced a variety of motions throughouttheir history. The attributes will alsoinclude the observational basis for theinterpretation (e.g., seismic reflection,surface geology, gravity, etc.) and theconfidence of the interpretation on, forinstance, a scale of 1-5.

Gravity and magnetic data canhelp interpret areas where seismic dataare insufficient. Using the GIS, it iseasy to extract profiles or regions to bemodeled, export them to a potential-field modeling package, and integratethe results back to the GIS when themodeling is complete. Aspecific exam-ple would be the construction of aninterface to export surfaces, well data,and other constraints to a 3-D GISmodeling package. (See “Suggestionsfor further reading” for more infor-mation.)

Spatial analysis is one of the keystrengths of GIS. For instance, wecould perform a spatial comparisonbetween seismicity and mapped faultlocations to assess the possible activ-ity along a fault or group of faults.Unfortunately, this would require bet-ter earthquake locations than currently available. These may be pos-sible in the near future with improvedoperation of the new Syrian seismicnetwork.

Tectonic model. The final goal of ourinterpretations is construction of a tec-tonic model for the Phanerozoic evo-lution of the northern ArabianPlatform. Figure 6 shows a prelimi-nary version of such a model andFigure 7 the corresponding timing ofevents. Naturally, as our analyses con-tinue, this model will become morerobust and refined. This is a relativelyeasy task if all information and pro-cessing is within the GIS platform.

The isopachs in Figure 6 have beendrawn within the GIS from well data,some seismic reflection constraints,and other information from publishedliterature for areas outside Syria. Whileschematic in Figure 6, in the final con-struction of this model, the “active tec-tonic elements” in each time frame willbe based on the attributes of fault activ-ity within the GIS; i.e., faults attrib-uted with normal fault movement inthe Late Cretaceous will be shown

with the appropriate symbols in theLate Cretaceous time frame.

In summary, our model shows thatepisodes of tectonic deformation expe-rienced by the Arabian Platform havebeen caused by activity on nearbyplate boundaries. The forces resultingfrom these activities are manifest inthe repeated reactivation of tectoniczones within the platform. These“weak” tectonic zones are possiblysuture zones relic from the Pan Africanaccretion of the Arabian plate and stillform the locus of deformation and tec-tonic activity in Syria today. Separatingthese tectonic zones are large stableareas that have experienced littledeformation. Examination of Figure 5,for instance, shows the concentrationof structures within the tectonic zonesas defined in Figure 1, with limiteddeformation outside of these zones.Attributes of the faults clearly indicaterepeated reactivation in various sensesthroughout time.

Although Early Paleozoic defor-mation is poorly represented in theGIS owing to data-quality issues, anunconformity of Late Silurian andDevonian age is observed regionallyin well data. This is possibly related tothe formation of the PaleoTethysOcean. Subsidence curves constructedfrom well and seismic data indicatethat since the Carboniferous, the north-ern Arabian Platform underwent aseries of extensional periods associ-ated with subsidence and rifting,

resulting in large sedimentary accu-mulations. The Palmyride/Sinjartrough formed during Carboniferoustime, possibly as a result of stress reor-ganizations on the Laurussian mar-gins. The more significant rifting inthe Palmyrides in Permian time wascoincident with the separation of theCimmerian block from Gondwana.This separation eventually led to thedemise of the PaleoTethys Ocean, asthe NeoTethys formed through accre-tion off the northeastern margin ofGondwana (Figure 6a). The formationand expansion of the NeoTethys in thePermian and Early Mesozoic was con-current with the early stages of riftingin the Eastern Mediterranean and rift-ing in the Palmyrides. The growth ofthe Mediterranean Basin andNeoTethys continued episodicallythroughout the Early Mesozoic,accompanied by rifting and subsi-dence in the Palmyride and Sinjarareas (Figures 6b and 7), that were pos-sibly an aulacogen at this time.

Consumption of the NeoTethysstarted in the mid-Mesozoic and bymid-Cretaceous time a subductionzone had formed near the northeast-ern margin of Arabia (Figure 6c).Widespread extension in the Sinjar andEuphrates areas in the latestCretaceous is evidenced by the largenumbers of normal faults there activeat that time (Figure 5). This extensionis thought to be a consequence of reor-ganized subduction in the NeoTethys

0000 THE LEADING EDGE FEBRUARY 2000 FEBRUARY 2000 THE LEADING EDGE 18

Figure 7. Summary chart showing the relative timing in the geologic evolu-tion of the eastern Mediterranean in general and Syria in particular. Notethe penecontemporaneous nature of many of the tectonic events through-out the region.

(Figure 6d), while abrupt cessation ofthis extension at the end of theCretaceous is probably a consequenceof continental collision with the trench(Figure 6e). Significant reverse motionon the older normal faults is firstobserved in western areas aroundMiddle Miocene time, progressingeastward, and leading to inversion innortheastern Syria in Pliocene time(Figure 6f). This sequence can beclosely correlated with the terminalclosure of the NeoTethys and suturingof Arabia to Eurasia, in concert withthe opening of the Red Sea.

Sedimentary analysis and hydrocar-bon exploration. The work discussedhere is mostly concerned with struc-tures and tectonics. A more completeintegration of sedimentalogic andstratigraphic data could better definetectonic episodes and paleogeographyby contributing information aboutwater depth, source areas, and so forth.Detailed field observations in thePalmyride fold and thrust belt and theSyria Coastal Ranges, in addition tolarge numbers of subsurface litholog-ical descriptions from drilling records,make such a synthesis possible.

We are currently integrating recentoutcrop stratigraphic observations intothe GIS. These additional data allowfor more accurate production ofisopachs, facies maps, paleogeo-graphic reconstructions, and basinanalyses that we eventually hope tocomplete for all areas of Syria. TheMesozoic section is the first target ofthis study owing to its overwhelmingeconomic importance in the northernArabian Platform.

Syria has a long history of hydro-carbon production. The currently pro-duced 550 000 b/d come from three ofthe tectonic zones that we have dis-cussed—the Palmyride fold and thrustbelt, northeast Syria, and, most impor-tantly, the Euphrates fault system. Thetectonic episodes that we have dis-cussed have been critical to the for-mation of these hydrocarbon reserves.For example, the phase of LateCretaceous extension in eastern Syriawas when many of the source rocks forthat area were deposited. The subsi-dence associated with this and othertectonic events also permitted the bur-ial of the source rocks deep enough forhydrocarbon generation. Additionally,the faulting was instrumental in theformation of many traps. The periodsof tectonic compression experiencedin Cenozoic time throughout Syriahave been critical to trap formation inthe northeast area, and for some

Palmyride fields.Our continuing effort to map the

structure and tectonic history of Syriawill help to focus future hydrocarbonexploration efforts in the country.Furthermore, with the completion ofour planned work to map the sedi-mentary and stratigraphic character-istics of the Syrian Mesozoic section,the benefit of this work for hydrocar-bon exploration will become moreapparent. The interrogation capabili-ties of the GIS excel when used inhydrocarbon exploration. When all thecoverages are built, it will be simple toselect, for example, all areas whereLower Cretaceous sandstone is morethan 20-m thick, adjacent to faulting,and proximal to a potentially maturesource rock. The wells within suchareas could then be evaluated for theirexploration successes, and undrilledareas could be identified. Once thenecessary coverages are in place, thiscould all be performed in a matter ofminutes.

Conclusions. GIS is an essential component in our current tectonicmapping of the northern Arabian Plat-form. Data manipulation, visualiza-tions, and integrated interpretations,previously prohibitively time-con-suming or impossible, are quick andeasy using this technology. This allowsfaster, more accurate analysis, and bet-ter-quality, highly accessible, end prod-ucts. The maps and models we arecreating are important for continuedhydrocarbon exploration within Syriaand surrounding areas.

This work is still in progress, andsome ideas discussed in this article arenot yet implemented. Once ourregional mapping of Syria is complete,we envisage Internet or CD-ROM pub-lication, in addition to paper distribu-tion, to allow digital distribution ofvarious coverages. See http://atlas.geo.cornell.edu/syria/welcome.html forupdates regarding this project.

Suggestions for further reading.“Middle East tectonics: Applications ofGIS” by Seber et al. (GSA Today, 1997)introduces some of the more extensiveWeb-based GIS development at CornellUniversity; also see http://atlas.geo.cornell.edu. “Geophysical and geological data-bases and CTBT monitoring: A casestudy of the Middle East” by Barazangiet al. (in Monitoring a CTBT, 1996) fur-ther details GIS activities at Cornell.“The petroliferous Euphrates graben sys-tem, Southeast Syria” by Litak et al.(AAPG Bulletin, 1998) and “Tectonic evo-lution of Northeast Syria: Regional tec-

tonics and hydrocarbon implications”by Brew et al. (GeoArabia, 1999) describedetailed tectonic research conducted atCornell. “Interactive visualization andmodification of 3-D models using GISfunctions” by Schmidt and Götze(Physical Chemistry of the Earth, 1998)describes the use of 3-D geoscience mod-eling, especially in respect to potential-field data. LE

Acknowledgments: The Cornell authors thankSyrian Petroleum Company (SPC) for the gen-erous provision of much of the data describedhere. Thanks to ARCO and Conoco for finan-cial support at Cornell, and to many colleaguesin Ithaca and Damascus for help and sugges-tions. In particular we thank Dogan Seber forhis extensive efforts in GIS development andresearch at Cornell.

Graham Brew is finishing his Ph.D. in geo-physics at Cornell University under the direc-tion of Muawia Barazangi. His research withthe Cornell Syria Project is focused on the inter-pretation of Syrian tectonic history throughgeophysical analysis. He obtained a B.Sc. in geo-physics from University College, London, andworked for a short time with RTZ Mining andExploration in Santiago, Chile.

Muawia Barazangi is a professor in theDepartment of Geological Sciences at Cornell,associate director of the Institute for the Studyof the Continents (INSTOC), and leader of theCornell Syria Project. His academic backgroundincludes a B.S. degree in physics and geologyfrom Damascus University, Syria, an M.S.degree in geophysics from the University ofMinnesota, and a Ph.D. in seismology fromColumbia University’s Lamont-Doherty EarthObservatory. His research history includesglobal tectonics, seismotectonics of continentalcollision zones, intracontinental mountain belts,and Middle East tectonics.

Tarif Sawaf currently works for the Al-Furatoil company in Syria while seconded from theSyrian Petroleum Company. Previously he wasa senior geologist and associate head of theregional mapping department for the SyrianPetroleum Company. Before that he worked forseven years with Sonatrach, Algeria. His aca-demic background includes a B.S. degree inapplied geology from Damascus University,Syria.

Khaled Al-Maleh is professor in the GeologyDepartment of Damascus University. Hiscareer has included time as head of GeologyDepartment, Damascus University, head of theDivision of Nuclear Ores Exploration with theSyrian Atomic Energy Commission, and gen-eral director of the Syrian GeneralEstablishment of Geology and MineralResources. He has published widely in French,Arabic, and English, and is currently focusedon issues of the Mesozoic development of Syria,especially in relation to phosphate deposition.

Corresponding author: G. Brew, brew@geology.cornell.edu

182 THE LEADING EDGE FEBRUARY 2000 JANUARY 2000 THE LEADING EDGE 00