201757

Upload: julian-boyer

Post on 04-Jun-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/13/2019 201757

    1/17

    Hindawi Publishing CorporationJournal o Geological ResearchVolume , Article ID ,pageshttp://dx.doi.org/.//

    Research ArticleRemote Sensing and Geographic Information System forFault Segments Mapping a Study from Taiz Area, Yemen

    Anwar Abdullah, Shawki Nassr, and Abdoh Ghaleeb

    Geology Department, Faculty o Applied Science, aiz University, aiz, Yemen

    Correspondence should be addressed to Anwar Abdullah; [email protected]

    Received June ; Revised September ; Accepted September

    Academic Editor: Karoly Nemeth

    Copyright Anwar Abdullah et al. Tis is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

    GIS and remote sensing data or allowing detection o structural eatures, such as aults, offer opportunities or improving mappingand identiying the areas that are likely to be locations o aulting areas. Landsat EM- satellite data images were used and band-was oundas the most suitable band or lineament delineation, based on theability to identiy geological eatures. Four contributingactors, namely, drainage patterns, aults (previously mapped), lineaments, and lithological contacts layers, were parameters usedin this study to produce a ault potential prediction map using the overlay model techniques. Te potential map(ault susceptibilitymap) classies the study area into ve potential zones, namely, very low, low, moderate, high, and very high potential. Te areascovered by moderate to the highest potential zones were considered as ault segments (ault lines) in the area. Te comparison o

    the potential map and the published ault map by using GIS matching techniques shows that ault segments (ault lines) in thepotential map were not properly identied in the study area. Te correlation between ault segments and aults data collected romeld work stations shows that there were aultsegments which may representnew aults in thearea being identied. Te presenceo these aults is not known rom the literature; this leads to updating and revising o existing geological map o the study area.

    1. Introduction

    Faults are weakness zones in the brittle parto the lithosphere,along which movement can take place in response to inducedstresses. When aults undergo displacement, depending ongeological and structural conditions, strain markers can beormed on the ault surace []. Te presence o aults in any

    area is based on displacement o rock layers. But also, mosto aults are represented by some geological eatures such asdrainage patterns, lineaments (linear eatures), and litholog-ical contacts between rock units within the rocks o the area.Te presence o aults may be indicated by these geologicaleatures (actors).

    Te term lineament was rst introduced by [, ] whorecognized the existence o linear geomorphic eatures andinterpreted them as surace expressions o zones o weaknessor structural displacement o the earths crust. Lineaments arelinear eatures on the Earths surace, usually related to thesubsurace phenomena. Generally, lineaments are related tolarge ractures and aults where their orientation and number

    give an idea o racture pattern o rocks []. In the recentyears, the lineaments have been dened as natural crustalstructures that may represent a zone o structural weakness[].

    Te drainage system, which develops in an area, is strictlydependent on the slope, the nature, and attitude o bedrockand on the regional and local racture pattern []. Most

    stream networks are adapted to regional slope and geologicalstructures, picking out the main ractures in the underlyingrocks [].

    Te contact between two lithologies can also appear as alinear eature. Tis contact may appear as a change in drai-nage pattern across the structural eatures [] orthe two unitsmay have different spectral properties [].

    Based on the denitions o lineaments, drainage patterns(including pattern the length, spatial distributions), and thelithological contacts between different rock types, the aultscould be mapped in the study area. Tese geological eaturesmostly resemble a ault lines in the area. Te most importanteatures in the area are the presence o drainage lines patterns

  • 8/13/2019 201757

    2/17

    Journal o Geological Research

    and ractures. Lineaments, drainages, lithological contacts,and previous ault lines data are important data and used inthis research or ault segments mapping using GIS technique.

    With the advento remotesensing andcomputer technol-ogy in the geosciences, geological investigation and interpre-tation have entered a new era. Remote sensing technology is

    very efficient or collecting data. Computer technology, suchas computer-based geographic inormation system (GIS),supplies a different method or data storage, integration,analysis, and display. Te combination o remote sensingand GIS provides an optimum system or various geologicalinvestigations such as ault mapping [].

    Several studies have been carried out to enhance geolog-ical knowledge and revise existing geological maps by usingoptical remotely sensed data and discovered new aults thatwere previously unmapped []. Furthermore, authors in[, ] conclude by suggesting that satellite images and/orgeographic inormation systems (GIS) are useul or tectonicmapping and identiying previously unmapped aults.

    Te benet o integrating different data and techniques instructural eatures such as aults identication and analysismade possible using remote sensing and GIS techniques isthe ability to identiy aults based on their characteristics.Creation and classication o various types o geological ea-tures can be carried based on the purpose o the study. Oncethe identication, preparation, and processes are complete,geographic inormation systems (GIS) unctionality, such as

    vector and raster spatial analysis and overlay, can be employedor structural mapping and analysis using powerul sofwareprograms [].

    Generally, Yemen is located in the southwestern part oArabian Peninsula, and it is affected by an active rifing zone

    o the Gul o Eden and the Red Seaopening. Tis active (rif-ing) zone can cause the reactivation o the old ault systemsand may create new aulting lines in the area. Te active aults(such as, rock mass movements or sliding on the ault planesand bedrock lithologies cracking) may cause severe damagesin some places, especially in urban or settlements areas. Forexample, bedrock cracking and rock mass sliding happenedin the and in some places in aiz city caused severedamages, and many houses, roads, and other inrastructureswere destroyed.

    Te previous works (data) such as geological and struc-tural maps are very old data, showing a ew o ault lines inthe studyarea. For this purpose, mapping o aults maylead to

    updating these data. In this paper, the using o remotesensing(data processing) and GIS (modeling) techniques could behelpul tools or mapping o aults, where the different geo-logical eatures (actors) are taken in the considerations. Teresults o the work could be considered rom the governmentor the people beore or during the different constructionprojects, to increase the saety, especially in the urban areas.

    Te purpose o this study was to test the integrationbetween remote sensing and geographic inormation system(GIS) or detecting the ault segments (ault lines) over thestudy area and to investigate the ability o this method ingiving real results compared to the previous data with respectto the ground truth (eld work).

    Te study area located in the southwest o Yemen has an

    area o km2. It includes aiz city and extends betweenRubayi and Al-Hawban areas (Figure ).

    According to GSY [] classication, the rocks in thestudy area are divided into two groups as ollows.

    () ertiary (granite body intrusive).

    () ertiary (volcanic rocks such as basalt, rhyolite,dacite, ignimbrite, and ashow deposits).

    Te geological map o the study area (Figure ) was dig-itized rom the geological sheet map o aiz with scale : ,.

    2. Materials and Methods

    Various data were used or ault segment susceptibility map-ping (potential ault zones mapping) in the study area. Var-ious image processing and enhancement techniques wereapplied on different remote sensing data including the

    Landsat EM- (Enhanced Tematic Mapper) satellite imagesand SRM (Te Shuttle Radar opography Mission) to getmaximum inormation by using different approaches oextraction methods. Te other data used in this study weretopographic maps, geological maps, and eld data.

    During conducting this study, many different sofwarepackages were used since there was no single sofware thatwould process all steps in the analyses. Tereore, the diversesofware used or the analysis in the current study is PCIGeomatica (version .), ERDAS (version .), and Ilwissofware (version .).

    .. Data Processing. According to this paper the linear ea-tures delineation was based on decision o the most appro-priate data, such as drainage patterns, lineament, ault (previ-ously mapped), and lithological contact layers or mappingo ault segments using GIS. Many o researchers such as[] have utilized lineaments, drainage patterns, aults,and lithology actors as important actors to measure andmap the susceptibility o geological hazard. Tereore, theinput attributes o the structural eatures o an area shouldbe studied as a rst consideration.

    Te previous data about the study area including geo-logical, structural, and topographic maps are digitized. oprepare these maps, rst, the maps are converted into a digitalormat by using the scanner. Tere are some data which were

    extracted rom data in digital ormats, such as Landsat EMsatellite images and SRM.

    Generally, our digital layers (maps) such as drainage pat-terns, ault (previously mapped), lineament, and lithologicalcontact were prepared and converted into secondary datasuch as drainage buffer, ault buffer, lineament buffer, andlithological contacts buffer layers.

    Tese layers were converted into slicing layers with veclasses. Each class was determined as m. Tis is becausethese layers were generated rom different data (EM, SRM,and previous maps) with different spatial resolutions andscales. Due to this, these layers should be enhanced andresampled (rescaled) to be suitable in resolution or GIS

  • 8/13/2019 201757

    3/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE

    N

    06

    04

    02

    00

    1508000mN

    1498000mN

    06

    04

    02

    00

    1508000mN

    1498000mN

    0.5 0 0.5 1 1.5 2(km)

    F : Location o the study area.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    Ignimbrite and ashflow deposits

    Rhyolite and dacite

    Basalt

    Granite

    Faults

    0.5 0 0.5 1 1.5 2(km)

    N

    F : Geological map o the study area [].

  • 8/13/2019 201757

    4/17

    Journal o Geological Research

    Data

    SRTMLandsat

    ETM-7

    Geometriccorrection

    Geometriccorrection

    Enhancement Enhancement

    Contourlines

    Bandselection

    DEM Filtering

    Digitizing

    Geologicalmaps

    Fault lines Contacts

    Drainageextraction

    Binaryimage

    Buffering

    Buffering

    Buffering Buffering

    Lineament

    Slicing

    Slicing

    Slicing Slicing

    Weightlayer

    Weight

    layer

    Weightlayer

    Weightlayer

    GISmodel

    Final fault map

    Evaluation

    Susceptibilitymap

    Fault segment

    Fault data

    Field data

    map

    F : Chart o the methodology in this study.

    model. Te resolution was selected as m or this processbased on the ability o satellite image o m spatial reso-lution to identiy the geological eatures in the area. Ten,

    slicing layers were converted into weight layers (thematiclayers). Tese procedures are illustrated inFigure .

    Arbitrary classications are still common; however, themain classication approaches are ranking, natural breaks,equal interval classes, equal area classes, and mean valueand standard deviation intervals [, ]. Hence, theclassication determines the spatial distribution o bufferingzones or susceptibility (classiying susceptibility) based onequal area classes.

    Classes (ve classes o each weight layer) and their weightvalues are given inable . Each class has a weight to expressthe contribution to the occurrence o ault segments. Teweight value was assigned to be between to (able ).

    : Parameters and the weight values are given to differentactors in the study area.

    Classes Weight value

    m

    Tese weight layers (maps) were input into GIS model.Te resultant map (susceptibility map) was classied intodifferent zones, very low, low, moderate, high, and very highault susceptibility zones or equal area classes using rankingapproach.

  • 8/13/2019 201757

    5/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE N

    0.5 0 0.5 1 1.5 2(km)

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    F : Original band- o Landsat EM-.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE N

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Band- o Landsat EM- afer contrast enhancement.

    ... Lineament Factor. Lineaments are related to large struc-tural ractures [] and it may represent zones o weakness

    []. Also, the presence o tectonic structures such as lineareatures could be important actor in geological hazardoccurrences [,].

    Filtering is useul in image enhancement. An edge-enhancing lter can be used to highlight any changes ogradient withinthe image eatures,such as structural lines [,,]. Tey generated lineament maps and determinedseveral signicant structural eatures.

    In the recent years, the lineaments have been denedas natural crustal structures that may represent a zone ostructural weakness []. Landsat EM- satellite data wereused and the rst step was to select the band that shouldbe used or lineament extraction []. Visual inspection o

    the individual bands was carried out, based on the abilityto identiy eatures, and band- (.. m) (SWIR) was

    selected as shown inFigure . And it was stretched linearlyto output range rom to (Figure ), because it is theleast affected band by the scattering, through the travel patho the long wavelength in the atmosphere. Tereore, it showsa good contrast and better display o geological eaturescompared to other bands.

    Te second step was to select the lter type. For thispurpose, different types o lters are tested such as by sobel kernel lter, by edge kernel lter, and by edgekernel lter. Te by and by edge enhancement ltersgive thicker and less linear areas afer threshold application;sobel results in thinner and more linear zones that give thebest result compared with other two lters.

  • 8/13/2019 201757

    6/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Treshold image produced by combining our ltered images.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE N

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Binary image produced by using sobel lters.

    : Sobel kernel lters in our main directions.

    Four main principal directions

    NW-SE E-W NW-SE N-S

    Te sobel lter was the best which convolved over band-in our principal directions and the ltered image was addedback to the original image. Te sobel lters in our principaldirections are given inable .

    Ten, the our enhanced ltered images combinedtogether to produce one threshold ltered image (Figure ),and this image was converted to binary image by applying

    adaptive thresholding. Tis adaptive threshold used to createbinary images is controlled by two parameters, and the suit-able values o adaptive thresholding parameters are windows

    size o , and this value is equal to thinning o pixels and themean offset o . Te nal binary image o the study area isshown inFigure .

    Te lineaments detected during the interpretation pro-cess were digitized directly on the image on the screen, andthe nal lineament map was recorded and stored in vectorles (Figure ).

    Tere are important points which need to be mentionedand should be ollowed in relation to the above procedure orlineament mapping. First, care is needed in the interpretationo the result o such analysis to exclude man-made eatures;thereore, the nal map was screened to remove lineamentsrelated to cultural eatures such as roads, canals, and eld

  • 8/13/2019 201757

    7/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Lineaments extracted rom binary image.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mE N 5.0

    4.2

    3.4

    2.6

    1.8

    1.0

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Lineament weight layer.

    plantation boundaries by comparing it with the topographicmaps and geological map o the area. All procedures men-

    tioned above have been ollowed during trace lineamentsrom binary image.

    Te lineament map was prepared and converted intolineament buffer layer and then converted into slicing layerwith ve classes. Finally, this layer was converted into weightslayer (Figure ). Te weight value was assigned to be between and (able ). And, this layer was ready to be input intoGIS modeling.

    ... Drainage Patterns Factor. Te drainage pattern isapparently being controlled by structure and lithology in thestudyarea.Te lithologicvariation hasgiven a rise to differentdrainage patterns. For example, radial and dendritic drainage

    patterns are developed over granitic rocks. Moreover, themost important eature in the area is the presence o drainage

    lines patterns and ault lines. It is clearly to see that there isa good relationship between these ault lines and drainagepattern system distribution especially with third and ourthriver orders in the area.

    Geological eatures are any alignment o eatures onsatellite images such as the various types recognized includ-ing topographic, drainage, vegetative, and color alignments.Digital elevation models (DEMs) are very useul in aidingthe classication o landorms or many purposes such asrecognition o drainage patterns. Extraction o topographiceature inormation rom DEMs has become increasinglypopular in structural analysis []. Digital elevation mod-els (DEMs) data were used to trace tectonic eatures and

  • 8/13/2019 201757

    8/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    2600m

    F : Digital elevation model (DEM) o the study area.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Drainage patterns map showing the nature o drainage system in the area.

    mapping geologically and topographically dened structuresin many areas []. Te Shuttle Radar opography Mission(SRM) data were also used to trace drainage pattern systemand mapping structures eatures []. Te extraction onegative system lines (ault lines) rom DEM data mostlyresembles drainage lines patterns [].

    SRM radar images with m resolution were used toobtain contour line data. Tis contour data was used to createthe DEM with m resolution o the study area. Te inter-polation and resampling processing techniques were used toconvert the contour line data (polygon) into m grid layerand the grid layer was converted into a DEM with m spatialresolutions (Figure ).

    Tis DEMs data was used or automatic extraction odrainage patterns over the study area using drainage basinrom elevation data unction (DRAIN) o PCI Geomaticasofware version .. Te drainage basin rom elevation dataunction (DRAIN) locates the drainage networks (river orstream) within the DEM. Te output data channel will appearas lines representing the rivers and streams [].

    Te result o the drainage pattern map using automaticextraction technique is shown in Figure . Tis techniquehelps us to extract the drainage patterns over the settlementarea, and the topographic maps were used as reerence data.

    Tis map was converted into vector layer. Te drainagepattern layer (vector layer) was converted into drainage buffer

  • 8/13/2019 201757

    9/17

    Journal o Geological Research

    0.5 0 0.5 1 1.5 2(km)

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN 5.0

    4.2

    3.4

    2.6

    1.8

    1.0

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Drainage weight layer.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Lithological contacts layer [].

    layer and then converted into slicing layer with ve classes.

    Tis map was converted into weights layer (Figure ). Teweight value was assignedto be between and (able ).Tislayer was ready to be input into GIS modeling.

    ... Lithological Contact Factor. Lithological contacts wereound in the eld, such as bedding planes and lithologicalcontacts between differentrock types. Te lithological contactlayer o the study area (Figure ) was digitized rom thegeological sheet map o aiz with scale : ,. Telithological contact layer (vector layer) was converted intolithological buffer layer and then converted into slicing layerwith ve classes. And this map was converted into weightslayer(Figure ). Te weightvaluewas assigned to be between

    and (able ). Tis map was ready to be input into GIS

    modeling.

    ... Fault Factor. Geological structures are patterns such asaults, olds, joints, and others ound in the rocks resultingrom natural deormation []. Te ault map o the studyarea (Figure ) was digitized rom the geological and thehydrogeological sheet maps o aizarea with scales : ,and : ,. Te rst sheet was prepared by geological sur-

    vey o Yemen (GSY) and the second was prepared by NationalWater Resources Authority, Dar El-Yemen Hydro consultant[]. Tis layer (ault layer) was converted into ault bufferlayer and then converted into slicing layer with ve classes.And this map was converted into weights layer (Figure ).

  • 8/13/2019 201757

    10/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    5.0

    4.2

    3.4

    2.6

    1.8

    1.0

    0.5 0 0.5 1 1.5 2(km)

    F : Lithological contacts weight layer.

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Fault map o the study area [,].

    Te weight value was assigned to be between and (able ).Tis layer was ready to be input into GIS modeling.

    .. GIS Model. Te major goal o this study is to producethe ault susceptibility and the ault segment maps in orderto get inormation about the possible locations o new aultsexisting in the area.

    Generally, our digital weight layers such as drainage pat-terns, ault (previously mapped), lineament, and lithologicalcontact layers were prepared rom different data. Te rststep o the GIS modeling involves the input data (weightlayers), and the second step includes selection o the suitableequation or this model to determine the susceptibility oaulting segments in the study area.

    In this study, the overlay layers model (spatial multilayersanalysis) was used preormed on these thematic maps with

    different weights by combining and aggregating these mapstogether with considering all different parameters. Te aultsusceptibility model produced with a modied linear equa-

    tion o drastic model [], using the ollowing equation:Potential Map= DRNW+ FLW+ LINW+ GCW, ()

    where, DRNW

    is the drainage weight layer, FLW

    the aultweight layer (previously mapped), LIN

    Wthe weightlayer, and

    LCW

    the lithological contact weight layer.Researchers such as [] classied the susceptibility

    into ve potential classes, very low (VL), low (L), medium(M), high (H), and very high (VH).

    Te resultant map (susceptibility map) was classied intodifferent zones, including very low, low, moderate, high, and

    very high ault susceptibility zones or equal area classes usingranking values.

  • 8/13/2019 201757

    11/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN 5.0

    4.2

    3.4

    2.6

    1.8

    1.0

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Fault weight layer.

    .. Evaluation o the Susceptibility Map. Te above-mention-ed technique was used to produce the ault segment suscep-tibility map rom the different data. Te result obtained romthis techniqueneedsto be evaluatedin some manner. Te rstevaluation was made with respect to the ault map (previousault map) using GIS overlay technique to determine wherethe ault segments and true aults are matched over the studyarea.

    During eldwork, structural data (strike/dip o aults)were collected rom stations in the study area. Tesedata were used or the ault segment map evaluation. Tese

    stations were located in the ault segments map to makewhether there is any correlation between the ault segmentsand the aults collected rom the eld.

    3. Results and Discussion

    Te nal susceptibility map o aulting (potential ault zones)was produced as a result o the integration o all parameterso weight layers (drainage, ault (previously mapped), linea-ment, and lithological contact). Tis map was classied intove zones, namely, very low potential zone (VL), low poten-tial zone (L), medium potential zone (M), high potentialzone (H), and very high potential zone (VH), as shown in

    Figure .Te potential zones in the ault susceptibility map show

    that the very low potential zone has a total area o .%,and the low potential zone has a total area o .%. Morethan hal o the study area are dominated by very low and lowpotential zones. And this area does not reect aults (aultsin the previous map), while the .% o moderate, .%o high, and .% o very high potential zones areas almostcover all the aults.

    Te evaluation o the susceptibility map has been com-puted by using GIS overlay capability techniques with theault map o the study area to determine where the potentialzones and aults are matched. In order to evaluate our

    susceptibility map or aulting, a comparison with the knownstructural ault lines has been done. Te result o thistechnique shows that all o the aults were matched with themoderate to high and very high potential zones (continuousareas rom moderate to high and very high).

    Tere were some o moderate to high and very highpotential areas in the susceptibility map that could not matchany ault lines in the output map o the ault map. Teseault segments identied by this technique may be suracereections o the new ault locations in the study area whichwere notidentied by previous works.Tis means that the use

    o this applied model worked successully to identiy lineareatures (e.g., ault lines) in the area o interest. Te resultssuggest an additional o new ault segment lines, whichwere not recognized by the ault map over the study area(Figure ).

    Moreover, the ault segments obtained rom this modelwere shorter in length compared with aults that wereobserved in the eld. For this reason, this paper attempted todetermine the accuracy o this technique or ault segmentsmapping and the relationship between these segments andground truth (aults) in terms o the length and the spatialdistribution in the study area.

    Based on the pattern o ault segments including the

    length, spatial distributions, and visual interpretation witha topographic map, a drainage pattern map, and threedimensions o DEMand Landsat EM, thetrue length o aultsegments could be recognized rom this analysis and couldbe implemented in order to assess the extraction accuracy oevery true ault length in the area.

    Validation o result was made with the geological truth(eld work) in the study area. During eldwork, many aultswere identied in different eld work outcrops. Strike/dipreadings o aults have been collected during the eld workrom different rock outcrops in the area. Based on the elddata, these linear eatures identied on Landsat imagery andDEM are interpreted to be aults (Figure ).

  • 8/13/2019 201757

    12/17

    Journal o Geological Research

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    16VH

    0.5 0 0.5 1 1.5 2(km)

    F :Susceptibilitymap oaultzones over thestudy area,classied into vezones, namely, verylow potential zone (green), low potentialzone (yellow), medium potential zone (blue), high potential zone (pink), and very high potential zone (cyan).

    386000 mE 88 90 92 94 96 98 400000 mE

    386000 mE 88 90 92 94 96 98 400000 mEN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    06

    04

    02

    00

    1508000 mN

    1498000 mN

    0.5 0 0.5 1 1.5 2(km)

    F : Te ault segments (aults lines) as a vector layer extracted rom the susceptibility map over the study area.

    Te eldwork stations were located in the ault segmentsmap to make whether there is any correlation between theault segments (aults lines identied by this technique) andthe aults data collected rom eld work stations. Tere wasgood correlation between a ew ault segments lines andaults, where locations came in the same or nearby thelocation o ault segments (aults lines). Tereore, these ault segments (aults lines) were identied as new aults(Figure ). Te structural reading (strike/dip) o aults rom eldwork stations are given inable .

    Tree ault systems were identied. Tese are the E-W,NW-SE, and NE-SW ault systems []. Te area was sub-

    jected to intensive tectonic activity which resulted in a series

    o grabens, horsts, andtilted ault blocks. Te tectonic activitydeormed the area beore the ertiary volcanic activity,especially through the NE-SW aultswhich run parallel to therif system o the Gul o Aden, and it is considered the oldestractures in the area. Te NW-SE aults are run parallel to theopening o the Red Sea and occurred in the area during theertiary volcanic activity. Te N-S and E-W ault system arethe youngest major aults in the area. Smaller aults (as well as

    joints) also ollow the same trends as the major aults. How-ever, more intensive aulting is in the NW-SF direction [].

    Conjugate aults completely dened the principal stressdirection. Strike and dip o the aults were plotted inequal-area projection o Schmidt net to obtain and calculate

  • 8/13/2019 201757

    13/17

    Journal o Geological Research

    : Structure reading o aults rom the eld.

    Location North East Strike/dip Name o ault Location North East Name o ault

    NE/ Normal Normal

    NE/ Normal Normal

    NE/ Reverse Reverse

    NE/

    Reverse Reverse NE/ Reverse Shear

    NE/ Shear Reverse

    NE/ Shear Reverse

    NE/ Reverse Normal

    NE/ Shear Normal

    NE/ Normal Shear

    NE/ Reverse Shear

    NE/ Normal Shear

    NE/ Normal Shear

    NE/ Normal Shear

    NE/ Normal Reverse

    NE/

    Normal Shear NE/ Reverse Shear

    NE/ Shear Reverse

    NE/ Shear Reverse

    NE/ Normal

    384973 401121

    384973 401121

    1508072

    1497024

    1508072

    1497024

    0 5(km)

    Faults

    N

    N (normal fault)

    R (reverse fault)

    S (shear fault)

    F : Te ault segments identied as aults in the study area showing true ault length and points o eld work stations, normal aultsegments indicated with red points, reverse ault segments indicated with green points, and strike slip ault segments indicated with bluepoints.

    the directions o principal stresses which acted on thoseaults.

    otally strike-dip readings o the conjugate aultsrom different rock units o the area were analyzed with theStereonet program (Stero Version .). Te stereographicplot was prepared (Figure ). According to the result o the

    conjugate aults analysis, compressional principal stress (1)is measured to be in the NE-SW direction.

    Tis analysis indicates that the N-S and E-W orientatedault sets are sets o shear ractures which could be typical orstrike slip aults. Te NW-SE and NE-SW orientated sets aresets o tensional and extensional ractures, which could be

  • 8/13/2019 201757

    14/17

    Journal o Geological Research

    0

    270

    180

    90

    F : Result o conjugate aults using stereographic analysis.

    typical or normal and reverse aults, respectively. Differenteld observations have also indicated that many damageswere ound or occurred along the NW-SE aults.

    4. Conclusions

    Tis study shows that remote sensing data and GIS techniqueare powerul tools in identiying geological structures such asault segments in the urban area such as the aiz city, Yemen.Te use o different types o data such as drainage patterns,aults (previously mapped), lineament, and lithological con-tacts with spatial resolution o m was combined through aprocess o integration and index overlay modeling techniqueor producing the susceptibility map (potential ault zones)o ault segments in the study area. GIS spatial overlaytechnique was used to determine the spatial relationships oall the criteria (actors) and subcriteria (classes) within layers(maps) to classiy and map the potential area o ault zoneso the study area. Te ault segments map shows that thereis good correlation and distribution between ault segmentsand aults (previous ault lines) in the study area. Tere are ault segments which may represent new aults in the area

    being identied. Te presence o these aults is not knownrom the literature. Tese aults could be active due to the rifsystem o the Gul o Eden and opening Red Sea. Active orreactivated aults can cause enormous damage to settlementor urban (houses or construction project) areas by rock massmovement along the ault suraces or bedrock cracks, damagewhich reers to loss o lie or injury, or damage to roads.Furthermore, geological survey o Yemen is requested toupdate the maps and create new ones using new data andtechniques and advise the people about the proper sitting ohouses and restricting settlement activities in danger areas.GISmodels can then yield more accurate estimateso the aultmapping using more data such as geophysical data.

    Acknowledgment

    Te authors would like to thank the Geology Department,Faculty o Applied Science, aiz University, aiz, Yemen orthe support to this research and staff or their valuable helpsand comments.

    References

    [] B. Dehandschutter, Study o the structural evolution o continen-tal basins in Altai, central Asia [Ph.D. thesis], .

    [] W. H. Hobbs, Lineaments o the Atlantic Border region,Geological Society, vol. , pp. , .

    [] W. H. Hobbs, Repeating patterns in the relie and in the struc-ture o the land,Geological Society, vol. , pp. , .

    [] L. E. Arlegui and M. A. Soriano, Characterizing lineamentsrom satellite images and eld studies in the central Ebro basin(NE Spain), International Journal o Remote Sensing, vol. , no., pp. , .

    [] R. . Walker, A remote sensing study o active olding andaulting in southern Kerman province, southeast Iran,Journalo Structural Geology, vol. , no. , pp. , .

    [] C. ravaglia and N. Dainelli, Groundwater Search by RemoteSensing: A Methodological Approach, Environment and NaturalResources, FAO, Rome, Italy, .

    [] M. Morisawa,Rivers, Longman, New York, NY, USA, .

    [] C. E. Brockmann, A. Fernandez, R. Ballon, and I. I. Claure,Analysis o geological structures based on landsat- images,in Remote Sensing Applications or Mineral Exploration, W.L. Smith, Ed., pp. , Dowden, Hutchinson and Ross,Strondsberg, Pa, USA, .

    [] P. . Nguyen and D. Ho, Multiple source data processing inremote sensing, inDigital Image Processing in Remote Sensing,

    J. P. Muller, Ed., pp. , aylor and Francis, Philadelphia,Pa, USA, .

    [] X. Chen, Application o remote sensing and GIS techniquesor environmental geologic investigation, northeast Iowa [Ph.D.thesis], University o Iowa, Iowa, Iowa, USA, .

    [] A. U.Akmanand K.uekci, Determination and characteriza-tion o ault systems and geomorphological eatures by RS andGIS techniques in the WSW part o urkey, in Proceedings othe th ISPRS Congress, pp. , Istanbul, urkey, .

    [] P. R. E. Guerra,Faulting evidence o isostatic uplif in the rinconmountains metamorphic core complex, an image processinganalysis [Ph.D. thesis], .

    [] M. A. Juhari and A. Ibrahim, Geological applications o

    Landsat M imagery: mapping and analysis o lineamentsin NW Peninsula Malaysia, in Proceedings o the th AsianConerence on Remote Sensing, pp. J--J--, Kuala Lumpur,Malaysia, .

    [] W. F. P. G. Micheal, Lineaments analysis South Florida region,aquier storage and recovery regional study, Draf echnicalMemorandum, Central and Southren Florida Project, ArmyCorps o Engineers, Jacksonville, Fla, USA, .

    [] I. D. Novak and N. Soulakellis, Identiying geomorphic ea-tures using LANDSA-/M data processing techniques onLesvos, Greece, Geomorphology, vol. , no. -, pp. ,.

    [] S. Solomon and W. Ghebreab, Lineament characterization andtheir tectonic signicance using Landsat M data and eld

  • 8/13/2019 201757

    15/17

    Journal o Geological Research

    studies in the central highlands o Eritrea, Journal o AricanEarth Sciences, vol. , no. , pp. , .

    [] K. S. Kavak and H. Cetin, A detailed geologic lineamentanalysis using landsat M data o Golmarmara/Manisa region,urkey,Online Journal o Earth Sciences, vol. , no. , pp. , .

    [] L. A. Rutty, Te basement racture pattern o sothern Ontario:a tectonic interpretation based on landsat M imagery, airphotosandeld data [M.S. thesis], National Libraryo Canada, Ottawa,Canada, .

    [] K. M. M. Elias, Multiple data set integration or structural andstratigraphic analysis o Oil and Gas bearing ormation usingGIS, in Proceedings o the Map India Conerence, Geology &Mineral Resource, .

    [] Geological Survey o Yemen (GSY), Geological Map o aiz Area(:, ), .

    [] S. Bai, J. Wang, G. N. Lu, P. G. Zhou, S. S. Hou, and S. N.Xu, GIS-based logistic regression or landslide susceptibilitymapping o the zhongxian segment in the three gorges area,China,Geomorphology, vol. , pp. , .

    [] R. L. Bates and J. A. Jackson, Glossary o Geology, AmericanGeological Institute, Alexandria, Va, USA, .

    [] M. L. Suzen and V. Doyuran, Data driven bivariate landslidesusceptibility assessment using geographical inormation sys-tems:a method and applicationto Asarsuyucatchment, urkey,Engineering Geology, vol. , no. -, pp. , .

    [] F. Guzzetti, A. Carrara, M. Cardinali, and P. Reichenbach,Landslide hazard evaluation: a review o current techniquesand their application in a multi-scale study, Central Italy,Geomorphology, vol. , no. , pp. , .

    [] C. F. Chung and A. G. Fabbri, Probabilistic prediction modelsor landslide hazard mapping, Photogrammetric Engineeringand Remote Sensing, vol. , no. , pp. , .

    [] F. C. Daiand C. F. Lee, Landslide characteristics and slope inst-

    ability modeling using GIS, Lantau Island, Hong Kong, Geo-morphology, vol. , no. -, pp. , .

    [] L. Ayalew, H. Yamagishi, and N. Ugawa, Landslide susceptibil-ity mapping using GIS-based weighted linear combination, thecase in sugawa area o Agano river, Niigata preecture, Japan,Landslides, vol. , pp. , .

    [] . B. Minor, J. A. Carter, M. M. Chesley, and R. B. Knowles, Anintegrated approach to groundwater exploration in developingcountries using GIS and remote sensing, in Proceedings othe International American Congress on Surveying and Map-ping/American Society or Photogrammetry and Remote Sensing(ACSM/ASPRS ), pp. , .

    [] S. Lee andK. Min, Statistical analysis o landslidesusceptibilityat Yongin, Korea, Environmental Geology, vol. , no. , pp.

    , .[] L. Donati and M. C. urrini, An objective method to rank

    the importance o the actors predisposing to landslides withthe GIS methodology: application to an area o the Apennines(Valnerina, Perugia, Italy),Engineering Geology, vol. , no. -, pp. , .

    [] A. Gunther, SLOPEMAP: programs or automated mapping ogeometrical and kinematical properties o hardrock hill slopes,Computers and Geosciences, vol. , no. , pp. , .

    [] S. A. Ali and S. Pirasteh, Geological applications o LandsatEnhanced Tematic Mapper (EM)dataand Geographic Inor-mation System (GIS): mapping and structural interpretation insouth-west Iran, Zagros Structural Belt, International Journalo Remote Sensing, vol. , no. , pp. , .

    [] G. Jordan, B. M. L. Meijninger, D. J. J. V. Hinsbergen, J. E.Meulenkamp, and P. M. V. Dijk, Extraction o morphotectoniceatures rom DEMs: development and applications or studyareas in Hungary and NW Greece, International Journal oApplied Earth Observation and Geoinormation, vol. , no. , pp., .

    [] A. Anwar, M. A. Juhari, and A. Ibrahim, Te extraction olineaments using slope image derived rom digital elevationmodel: case study o Sungai LembingMaran area, Malaysia,Journal o Applied Sciences Research, vol. , no. , pp. ,.

    [] K. Koike, S. Nagano, and M. Ohmi, Lineament analysis osatellite images using a Segment racing Algorithm (SA),Computers and Geosciences, vol. , no. , pp. , .

    [] A. Mah, G. R. aylor, P. Lennox, and L. Balia, Lineament anal-ysis o Landsat Tematic Mapper images, Northern erritory,Australia, Photogrammetric Engineering & Remote Sensing, vol., no. , pp. , .

    [] A. Anwar, M. A. Juhari, andA. Ibrahim, A comparison o land-sat M and SPO data or lineament mapping in Hulu Lepar

    area, Pahang, Malaysia, European Journal o Scientic Research,vol. , no. , pp. , .

    [] A. Masoud and K. Koike, ectonic architecture through Land-sat- EM+/SRM DEM-derived lineaments and relationshipto the hydrogeologic setting in Siwa region, NW Egypt,Journalo Arican Earth Sciences, vol. , no. -, pp. , .

    [] M. L. Suzen and V. oprak, Filtering o satellite images ingeological lineament analyses: an application to a ault zone inCentral urkey, International Journal o Remote Sensing, vol. ,no. , pp. , .

    [] A. Ganas, S. Pavlides, and V. Karastathis, DEM-based mor-phometry o range-ront escarpments in Attica, central Greece,and its relation to ault slip rates, Geomorphology, vol. , no.-, pp. , .

    [] S. Sarapirome, A. Surinkum, and P. Saksutthipong, Applicationo DEM data to geological interpretation: Tong Pha Phumarea, Tailand, inProceedings o the rd Asian Conerence onRemote Sensing (ACRS ), Kathmandu, Nepal, .

    [] G. Sarp and V. oprak, An Integrated Lineament Analysis romSatellite Images, inProceedings o the th Asian Conerence onRemote Sensing (ACRS ), Kuala Lumpur, Malaysia, .

    [] A. Anwar, M. A. Juhari, and A. Ibrahim, Automatic mappingo lineaments using shaded relie images derived rom digitalelevation model (DEMs) in the MaranSungi Lembing area,Malaysia, Electronic Journal o Geotechnical Engineering, vol. ,pp. , .

    [] P. C. I. Geomatica, PCI Geomatica Users Guide Version . ,Richmond Hill, Ontario, Canada, .

    [] A. Ibrahim and M. A. Juhari,Dictionary o Geological the Basicerms, Malaysia, National University o Malaysia, Selangor,Malaysia, .

    [] National Water Resources Authority, Hydrogeologic Map o aizArea (: , ), Dar El-Yemen Hydro Consultant, .

    [] L. Aller, . Bennett, J. H. Lehr, and R. J. Petty, DRASIC: astandard system or evaluating groundwater pollution potentialusing hydrogeologic settings, ech. Rep. EPA// /R.S., Kerr Enviromental Research Laboratory, EnviromentalProtection Agency, Ada, Okla, USA, .

    [] M. J. Crozier, Field Assessment o Slope Instability, in SlopeInstability, D. Brunsden and D. Prior, Eds., pp. , JohnWiley and Sons, New York, NY, USA, .

  • 8/13/2019 201757

    16/17

    Journal o Geological Research

    [] Q. Zaruba and V. Mencl,Landslides and Teir Control, Elsevier,Amsterdam, Te Netherlands, .

    [] I. Das, S. Sahoo, C. van Westen, A. Stein, and R. Hack, Land-slide susceptibility assessment using logistic regression and itscomparison with a rock mass classication system, along a roadsection in the northern Himalayas(India), Geomorphology,vol., no. , pp. , .

  • 8/13/2019 201757

    17/17

    Submit your manuscripts at

    http://www.hindawi.com