62 Three-dimensional Subsurface Modeling of Mineralization a Case Study From the Handeresi (Çanakkale, NW Turkey) Pb-Zn-Cu Deposit

8/10/2019 62 Three-dimensional Subsurface Modeling of Mineralization a Case Study From the Handeresi (anakkale, NW Tu

1/14

574

http://journals.tubitak.gov.tr/earth/

Turkish Journal of Earth Sciences urkish J Earth Sci(2013) 22: 574-587 BAKdoi:10.3906/yer-1206-1

Tree-dimensional subsurface modeling of mineralization: a case study from theHanderesi (anakkale, NW urkey) Pb-Zn-Cu deposit

Sinan AKISKA1,

*, brahim Snmez SAYILI2, Gkhan DEMRELA

3

1Department o Geological Engineering, Faculty o Engineering, Ankara University, andoan, Ankara, urkey

2Fe-Ni Mining Company, Balgat, Ankara, urkey

3Department o Geological Engineering, Faculty o Engineering, Aksaray University, Aksaray, urkey

* Correspondence: [email protected]

1. IntroductionTree-dimensional (3D) interpretation o subsuracecharacteristics has been used in the mining sector or along time. Beore the use o state-o-the-art computersofware, portrayal o the 3D eatures was done usingtwo-dimensional (2D) specialized maps, cross-sections,and ence diagrams. Currently, it is possible to construct3D subsurace models easily using 3D Geoscientific

Inormation Systems (3D GSIS), which have efficient data-management capabilities (Rahman 2007). In the last 2decades, the number o 3D subsurace modeling studieshas increased due to the use o computer sofware (e.g.,Renard & Courrioux 1994; de Kemp 2000; Xue et al.2004;Feltrin et al.2009; Ming et al.2010; Akska et al.2010b).High-definition 3D models are constructed using theinterpolation algorithms o those sofware programs; inaddition, the determination o underground mines andtheir conditions o ormation can be obtained. Becauseo connections between the study areas and the data thathave some differences in all 3 dimensions, 3D GSIS isimportant (Rahman 2007). Te application areas o 3DGSIS are: determining ore and oil deposits (e.g., Houlding

1992; Sims 1992; Feltrin et al. 2009; Wang et al. 2011),hydrogeological studies (e.g., urner 1992; Houlding1994), various civil engineering projects (e.g., zmutlu &Hack 1998; Veldkamp et al.2001; Elkadi & Huisman 2002;Rengers et al.2002; zmutlu & Hack 2003; Zhu et al.2003;Hack et al.2006, Bistacchi et al.2008), modeling structuralactors (Renard & Courrioux 1994; de Kemp 2000; Galeraet al.2003; Zanchi et al.2009), and establishing settlement

areas (e.g., Rahman 2007). Te main aim o 3D modelingo ore deposits is to determine the complex geological,structural, and mineralogical conditions in these areas andto detect the location o these deposits in the subsuraceenvironment. With the help o recent 3D subsuracemodeling studies, some inormation can be gained not onlyabout detecting ore locations but also about the ormationconditions o the deposits (e.g., Feltrin et al.2009).

Optimizing the subsurace data collected rom varioussources (boreholes, geophysical methods, well logs, etc.)could minimize the costs o many operations. Because othe complex spatial relationship existing in the subsurace

environment, regularly spaced boreholes and good-qualitydata are necessary to resolve this complexity (Hack et

Abstract: Te main goal o 3D modeling studies in the mining sector is to address the complex geological, mineralogical, and structuralactors in subsurace environments and detect the ore zone(s). In order to solve this complexity, use o quality data (e.g., a wide range oboreholes at regular intervals) is necessary. However, this situation is not always possible because o certain restrictions such as intensive

vegetation, high slope areas, and some economic constraints. At the same time, with the development o computer technology, theunused and/or insufficiently considered data need to be gathered and reviewed. Tis assessment may lead to the detection o potentialnew zone(s) and/or could prevent unnecessary costs. In this study, the target area that was chosen had inadequate and unusable data,and we used the data as effectively as possible. Te Handeresi area is located in the Biga Peninsula o northwestern urkey. In this area,the Pb-Zn-Cu occurrences take place in carbonate levels o metamorphic rocks or at the ractures and cracks o other metamorphicrocks. Te area is being explored actively now. In this study, using the borehole data, we attempted to model the subsurace o this areain 3D using commercial RockWorks2006 sofware. As a result, there were 3 ore zones that were seen intensively in this area. One othem indicates the area in which the adits are now operating. Te others could be new potential zones.

Key words: NW Anatolia, Biga Peninsula, inverse distance weighting, kriging, lead, zinc, copper, interpolation method

Received:05.06.2012 Accepted:19.12.2012 Published Online:13.06.2013 Printed:12.07.2013

Research Article


2/14

575

AKISKA et al. / Turkish J Earth Sci

al. 2006). However, this situation is not always possibledue to economic constraints and the difficulties o fieldconditions.

In urkey, some institutions such as the GeneralDirectorate o Mineral Research and Exploration o urkey

(MA), urkish Petroleum Corporation (PAO), GeneralDirectorate o State Hydraulic Works (DS), and othershave thousands o meters o borehole log data. Tesedata were interpreted using old technology, and much othe inormation is no longer used today. In act, some othese data were taken casually and could not be associatedwith the subsurace characteristics (especially due tothe technological deficiencies at that time) and/or couldnot be interpreted due to inadequacies in the number oboreholes in the study areas. With the development onew technology, these unused data need to be gatheredand reviewed. New ore zones can then be detected, and

unnecessary costs can be prevented by means o thereviewed data.

Te goals o this study were to determine the potentialPb-Zn ore zones in the subsurace environment o theHanderesiCu-Pb-Zn deposit by means o the surace andborehole geologic data, and to provide ocus on mining

operations in specific areas in spite o obstacles such asinsufficient borehole units, structural actors, and intensive

vegetation. In addition, it is hoped that this study will makea contribution to more detailed modeling studies.

2. Geological settingTe study area is located in the Biga Peninsula onorthwestern urkey. It is situated between the Edremit(Balkesir) and Yenice (anakkale) districts, and lies tothe south o Kazda Massi in the western section o theSakarya Zone (Figure 1). Tis zone is represented by Pre-Jurassic basement rocks that are deormed, and it includesmetamorphosed and unmetamorphosed Jurassic-ertiaryunits. Te area consists o Devonian (Okay et al. 1996)granodiorite rocks called amlk granodiorite, a Permo-riassic (Okay et al.1990) metamorphic sequence calledthe Karakaya Complex, and Oligo-Miocene (Krushensky

1976) granitoid and volcanic rocks. Te common rocks inthe metamorphic sequence are sericite-graphite schists,phyllites, and quartzites with metasandstone and marblelenses (Figure 2). Te Pre-Jurassic clues o the basementunits are strongly overprinted by Alpidic deormations(Okay et al.2006).

26 34 42

36

Strandja

zone

racanbasn sta

nbul

zone

Intrapontdesuture

Kazd

a

massf

Born

ovaflys

chzon

eTavanlzone

Afyonzone

Menderesmassf

Krehrmassf

Lycannappes

PamphlanSutu

re

BLACK SEA

AEGEA

NS

EA

Sakarya zone

N

W

S

E

Studyarea

ANKARA

ZMR

STANBUL

BLACK SEA

MEDITERRANEAN SEA0 200 km

Studyarea

ISPARTA

RKEYRAEGEANS

EA

Figure 1. Simplified tectonic map showing the location o the main ethyansutures and neighboring tectonic units in western urkey (afer Okay et al.1990;Harris et al.1994).


3/14

576


BB

+

+

+

+

+

+

HANDERES

YG

HDYU

HDK

7

22 5

20 6

23

11

27

26

9 8

24

13

14

17

16

25

12

12

10

15

3

18

19

21

4

A

A

B

B

a

Q

a

Pz

Pz

PzPz

Pz

Pz

Pzmk

Pzmk

Pzmk

Pzmk

Pzmk

Pzmd

Pzmd

Pzmd

Pzmd

Pzmd

Pzmd

Pzmk

Pzmk

0 100 200 m

N

W

S

E

METASANDSTONE

SCHIST

METADIABASE

VOLCANITES

ALLUVIUM

CROSS-SECTION LINE

FAULT-PROBABLE

FAULT

ACTIVE-DEACTIVEMINING

RIVER

ROAD

BOREHOLES

SYNCLINE-ANTICLINE

Q

a

Pzmd

Pzmk

Pz

8

A

A

35/518100

4402200

35/519100

4400800

Figure 2.Geologic map o the Handeresi area including the cross-section lines between the boreholes; coordinates are given inUM coordinate system (modified rom Ycelay 1976).


4/14

577


Some vein- and skarn-type lead, zinc, and copperdeposits are located in the Permo-riassic metamorphicsequence (Ycelay 1976; aatay 1980; etinkaya et al.1983; uan 1993; Akska 2010). Mineralized zones occurin carbonate levels o metamorphic rocks or at the ractures

and cracks o other metamorphic rocks. Te main oremineral paragenesis is galena, sphalerite, chalcopyrite,pyrite, arsenopyrite, and hematite assemblage, whilegangue minerals are grossularitic and andraditic garnets,manganierous hedenbergitic pyroxenes, epidote, quartz,and calcite (Akska 2010; Akska et al.2010a; Demirela etal.2010).

When the Handeresi Pb-Zn-Cu deposit was exploredin the early 1970s by the MA, 27 boreholes were drilledor mineral exploration. Te Handeresi deposit is one othe most important Pb-Zn-Cu occurrences in urkey, with

total mineral resources o 3.5 Mt at an average grade o 7%Pb, 4% Zn, and 3000 g/t Cu (Ycelay 1976). In this area,mining activities have been maintained or about 40 years.Te deposit is currently mined by Oreks Co. Ltd., whichhas produced ores rom 4 adits in this area.

3. MethodsAs pointed out previously, subsurace modeling studies arequite important in mining sectors, and detailed modelingstudies can be achieved as a result o developmentsin computer technology. Particularly, assigning aditdirections accurately in an underground mining area canreduce the various operating costs associated with mining.

Several sofware programs that are used or surace andsubsurace modeling include several spatial interpolationalgorithms such as nearest neighbors, inverse distanceweighting (IDW), kriging, and triangulated irregularnetwork-related interpolations (Li & Heap 2008).

Each sofware program has its own advantages anddisadvantages, but all o them have almost every one othe interpolation algorithms used in modeling studies(Rahman 2007). In this study, modeling was accomplishedwith commercial RockWorks2006 sofware, whichenables the use and capability o the interpretation inconjunction with all o the data in much less time. Tisprogram includes 2 main windows: Borehole ManagerandGeologic Utilities. Borehole Manager contains the boreholeprocedures such as entry, management, and analysis oborehole data. Geologic Utilities has mapping, gridding,

and contouring properties (Rahman 2007). In this study,Borehole Manager is used or subsurace modeling andGeologic Utilitiesis used or surace modeling (Figure 3).

In this study, the database is created using borehole,topographic, and geologic data, and a digital elevationmodel (DEM) is generated via topographic data. DEMis used especially in GIS applications constructing 3Dsurace modeling. A 3D topographic map is generatedby determining the unknown points rom certain pointsthrough various interpolation methods. Tat is whychoosing the right interpolation method is very importantor creating DEMs. Many researchers have revealed therelationship between DEM accuracy and the interpolation

GEOLOGICAL

MAP

TOPOGRAPHIC

MAP BOREHOLE DATA

GEOREFERENCED

GEOLOGICAL

MAPDEM

SURFACE MODELING SUBSURFACE MODELING

LITHOLOGY

GEOCHEMICAL

ANALYSES

(Pb%, Zn%, Cu%)

BOREHOLE

POINT MAPS

SOLID

MODELING

SOLID

MODELING

WITH CUT-OFF

GRADES

2D CROSS-

SECTIONS

3D BOREHOLE

DATA

SOURCE

PROCESSES

KRIGING

IDW IDWLITHO BLEND

Figure 3. Organigram o the modeling processes (modified rom Kaumann & Martin 2008) (the words in the parallelogramsindicate the interpolation methods).


5/14

578


technique (Zimmerman et al. 1999; Binh & Tuy 2008;and reerences therein). Fenck and Vajsblov (2006)investigated the accuracy o the DEM using the kriginginterpolation technique with different variogram modelsin the Morda-Harmonia area. As a result o this study, the

authors concluded that the most appropriate variogramwas a linear model. Chaplot et al. (2006) created DEMsusing various interpolation techniques (kriging, IDW,multiquadratic radial basis unction, and spline) inregions o France and Laos. Te authors concluded thatall interpolation techniques showed similar perormancein the regions with dense sample points, while IDWand kriging were better than the others in regions withlow density sample points. However, the study carriedout by Peralvo (2004) in 2 watersheds o the EasternAndean Cordillera o Ecuador showed a different result.According to this study, the IDW interpolation methodproduced the most incorrect DEM. In the evaluation othese studies reerred to by Binh and Tuy (2008), theauthors noted that the studies showed contradictoryresults due to the differences in technological applicationlevels, research methods, and the types o topography indifferent countries.In their study, Binh and Tuy (2008)created DEMs via 3 interpolation techniques in 4 differentareas in Vietnam using digital photogrammetry and totalstation/GPS research methods. As a result, regularizedspline interpolation is the most suitable algorithm inmountainous regions, while IDW or an ordinary kriging

interpolation algorithm with the exponential variogrammodel is recommended in hilly and flat regions (Binh &Tuy 2008). Whenall the data are evaluated together, eventhough some points are important while choosing theinterpolation method that creates the DEM, there are nospecific rules or choosing the interpolation algorithms.Nevertheless, considering the work done by Binh and Tuy(2008), because the area in this study includes flat and hillyareas, the kriging interpolation method is preerred orsurace modeling.

Kriging (Krige 1951; introduced by Matheron 1960) is

the generic name o generalized least-squares regressionalgorithms (Li & Heap 2008). Tis method is a well-known geostatistical interpolation method that weightsthe surrounding measured values to derive a predictionor an unmeasured location (Cressie 1990). Tis algorithmis an estimation process that determines the unknown

values using the known values and variograms. Krigingis considered the most reliable method or geological andmining applications (Rahman 2007). Te most importantadvantage o kriging, compared with other estimationmethods, is that the weights are determined via certainmathematical operations instead o randomly. Te data are

analyzed systematically and objectively; as a result o thisanalysis, weights that will be used in variogram unctions

are calculated (ercan & Sara 1998). Another advantageo this method is that it gives the error estimation viathe kriging variance. Te kriging variance does notdepend on the exact values o the data; it is a unctionbetween the numbers o data and the distances o data

(ercan 1996). Very close estimation o data generatedby the kriging interpolation algorithms to the real valuesdepends on the number o samples, the requency o data,and the degree o accuracy o the variogram model andparameters (Brooker 1986; Chaouai & Fytas 1991). Temethod creates variogram models o the data set thatrepresent the relation o the variance o the data pairs withdistance. Tis variogram indicates the extent o the spatialautocorrelations and the variogram models that couldbe isotropic or anisotropic, depending on the directional

variability o the data. Te unknown values are predictedbased on the variogram model (Cressie 1990). Mostrequently used in several variogram models are spherical,exponential, linear, and Gaussian models (Burrough &McDonnell 1998).

While doing subsurace modeling, RockWorks2006perorms solid modeling. Solid modeling is a grid processin 3 dimensions, which creates a cube rom regularlyspaced nodes derived rom irregularly spaced data. During3D modeling, the subsurace is divided into cells that havespecific dimensions called voxels, and the geologic unitsthat correspond to these cells orm the cubes. Each voxelcreated is identified by the corner points, called nodes.

Each node has an x, y, and z location coordinate, and a gvalue, which in this study is a geochemical analysis value.In this study, 2 different solid models are made. Te firstmodel is applied to ORE ZONE, shown in the boreholes.Te second model is applied to Pb%, Zn%, and Cu% valuesobtained rom these ore zones.

In the first model, RockWorks2006 uses a solidmodeling algorithm that is designed specifically tointerpolate lithologies in the boreholes. Using thisalgorithm, which is called litho blend, the subsuraceis separated into block diagrams and all lithologies are

modeled (RockWare 2006). In this study, all lithologicunits are modeled; however, only ORE ZONE is used orthe purpose o the study.

In the second model, in order to model the percentagedistribution o ore zones in the boreholes, Pb%, Zn%, andCu% values are modeled with 3D solid modeling.In thismodeling study, to generate the block diagrams in thesubsurace, the IDW interpolation method is preerred,which makes a distinction with respect to the similarityo degrees o the measured points. In other words, inthe estimation o the unknown points, it gives moreweight to the closest known points instead o the remoteones.IDW is very versatile and an easily understandableprogrammable method. In addition, it gives very accurate


6/14

579


results in wide-range data interpretation (Lam 1983). Temost important eature o this method is that it is able toquickly interpolate the scattered data in the regular gridsor the irregularly spaced data (Li & Heap 2008).

4. Geostatistical analysis of the surface dataTe number o x, y coordinates and z elevation data pointsin the area is 5292. Tese data are digitized rom thetopographic map rom Ycelay (1976). Te topographicmap has a 10-m contour interval o the study area. Teinormation about the survey method was not given byYcelay (1976).

As mentioned above, the modeling studies are carriedout using RockWorks2006 commercial sofware. However,or the geostatistical analysis, more comprehensivesofware is needed. Tereore, the geostatistical analysisis done with the Geostatistical Analyst ool in ArcGis9

(Johnston et al. 2001).Te changes depending on the distance o the difference

between regionalized variable values are revealed withthe variogram unction in geostatistics (ercan 1996).When the variogram is calculated in different ways, itsometimes exhibits different behaviors (Armstrong 1998).Anisotropy is used or calculating the directional effectsin the semivariogram model, which is made or suracecalculations. It is characteristic o a random process thatindicates higher autocorrelation in one direction thananother (Johnston et al.2001). In this study, the surace dataindicate anisotropy. Te range values are different while

the sill values are the same in the variograms calculatedin different directions in this study. Tis also shows thatthe surace has geometric anisotropy (Armstrong 1998).It can be seen that the major axis o the anisotropic ellipseis trended NE-SW (Figure 4a). Experimental variogramshave been calculated in 4 directions, which are N-S, E-W,NE-SW, and NW-SE. Te lag size is 100 m and the angletolerance is 45. Te experimental variogram has beenfitted by an exponential variogram model (Figure 4b)that represents the direction o maximum continuity 55rom the north. In the experimental variogram, the sill

value is 5392.6 m, the range is 1280.39 m, and the nuggetvalue is 10 m.

Neighborhood estimation, which defines a circle (orellipse) including the predicted values on unmeasuredpoints, is used to restrict the data (Johnston et al. 2001).

While interpolating each grid node, the search ellipsedefines the neighborhood o points to consider. Outsidethe search ellipse, the data points are not taken intoaccount (Fenck & Vajsblov 2006). In most cases, thesearch ellipse range and direction coincides with theanisotropy range and direction. At the same time, toprevent the tendency o particular directions, this circle(or ellipse) is divided into sectors. In this study, ordetermining the search ellipse, the anisotropy range anddirection are used automatically and the ellipse is dividedinto 4 sectors. Te maximum number o samples chosen is6 or neighborhood estimation.

Cross-validation is used to control all numbers o datapoints (5292 points) used in interpolation. Te graphicand table obtained afer the cross-validation analysis areshown in Figure 5 and able 1, respectively.

For perect prediction, the estimation errors should besymmetrically distributed, and linear regression o exact

values on estimated values should be close to a 45 line(Sara & ercan 1996). Te needed criteria or the bestcreated DEM were given by Johnston et al.(2001):

Standardized mean nearest to 0. Smallest root mean square (RMS) prediction error. Average standard error nearest to the RMS prediction

error. Standardized RMS prediction error nearest to 1.Both the predicted values are nearly the same as

measured values and the prediction error values indicatethe satisactory result o the interpolation (Figure 5; able1).

5. Tree-dimensional subsurface modeling ofmineralizationsTe study area covers 1.4 km2(1 1.4 km) and elevationranges rom 270 m to 520 m. Te surace has been divided

a) b)10-4

0 1 .62 3 .24 4 .86 6 .48 8 .1 9 .72 11.34 12.96

0.26

0.52

0.78

1.04

1.3

Dstance, h10-2

Figure 4.(a) Anisotropic ellipse showing NE-SW trend and the direction o the variogram,(b) experimental variogram in the direction o the major axis o the anisotropic ellipse.


7/14

580


into 10 10 10 m blocks (490,000 total voxels). In thenorthwestern and southeastern parts o the area, hilly

topography with gentle slopes is seen while the HanderesiRiver flows rom the northeast to southwest.In order to create surace modeling, the topographic

map (1/1000) o the study area (Ycelay 1976) is digitized,and x, y, and z values are entered into the Geologic Utilitiessection o RockWorks2006. Using these values, the sofwarecreates a grid-based file. While creating the grid file, thekriging method is used as an interpolation algorithm.One o the important eatures o RockWorks2006sofware is that it is able to choose the most appropriate

variogram that analyzes all the data automatically in thekriging interpolation method calculations. In this study,

the Exponential with nugget variogram determined inaccordance with the analysis o the sofware is preerred.Choosing the Exponential with nugget variogramautomatically shows that the results in this analysis arealso compatible with the results o geostatistical analysis.Te 3D topographic surace modeling is intersected with ageoreerenced geological map. Finally, the borehole pointmaps and the adits are done by drawing 3D boreholemultilogs (Figure 6).

Te subsurace has been divided into 5 5 5 m blocks.ORE ZONE applied to the litho blend algorithm and

Pb%, Zn%, and Cu% values applied to IDWalgorithm have4,010,151 and 2,154,921 total voxel values, respectively.

In order to create subsurace modeling, 27 borehole dataare taken rom Ycelay (1976). Te shallowest drilling is60 m (S-01) and the deepest drilling is 245.65 m. (S-13).otal drilling depth is 4239.15 m while the average drillingdepth is 157 m. Te ore zones are observed in 6 out o 27

boreholes (S-04, S-06, S-14, S-15, S-19, and S-21). In orderto determine Pb%, Zn%, and Cu% values in the ore zones,geochemical analysis was carried out according to themethods o Ycelay (1976). All o these values are enteredinto the RockWorks2006 sofware separately without anymodification. Te ORE ZONE, which is detected romdrilling cores, is modeled via solid modeling. Here, whilecreating the block diagrams, the sofware makes the solidmodels o the lithologies using the litho blendalgorithm(RockWare 2006).Tis algorithm is used to interpolate andextrapolate numeric values that represent ORE ZONEin the lithology class. Grid nodes between the boreholesare assigned a value that corresponds to the ORE ZONEsection in the lithology class and relative proximity oeach grid node to surrounding boreholes (Sweetkind et al.2010). Te modelis intersected with topographic suracemodeling. However, because o the insufficient number odrilled boreholes, the accuracy o the modeling o areasthat are outside o the drilled area (Figure 7) is arguable.

Using the Pb%, Zn%, and Cu% geochemical analysisresults, the model files are constructed separately usingthe IDW interpolation method. Te parameters o theIDW interpolation method and 3D grade results are

shown in able 2 and Figure 8, respectively. Te orezones determined in the model files, due to the existenceo ore zones in almost all boreholes, do not provide anyocus area. One o the aims o this study is to lead to moredetailed studies and to ocus the ore zone(s) into morerestricted areas. Determining o the area(s) in which Pb,Zn, and Cu mineralizations above the cut-off grades isthought to be ensured, as much as possible, close to thepurpose described above. For this purpose, using Pb%,Zn%, and Cu% values with the above cut-off grades inall ore zones creates a database in RockWorks2006. In

this sofware, this kind o subsurace data (such as thoserepresenting geochemistry, geotechnical measurements,etc.) is possible to model in 3D (I-data tool; RockWare2006). Using this tool, RockWorks2006 interpolates thedownhole interval-base data into a solid model. Solidmodeling is implemented separately or the chemicalanalysis belonging to each element (Pb.mod file or Pb%modeling, Zn.mod file or Zn% modeling, and Cu.modfile or Cu% modeling; Figure 9a).As mentioned above,the ore zones that exist in almost all the boreholes reflectthis modeling study, and large areas are detected or eachelement in the subsurace environment.Choosing a target

area is difficult when the results are considered together.Tat is why using the intersections o all element zones

R = 0.9999

250.00

300.00

350.00400.00

450.00

500.00

550.00

250.00 300.00 350.00 400.00 450.00 500.00 550.00

Pred

cte

dva

lues

Measured values

Surface data (krgng)

able 1. Te summary statistics o the prediction errors usingkriging interpolation with Exponential with nuggetvariogram.

Prediction errors

Samples 5292

Mean 0.007349

RMS 0.7163

Average standard error 1.6

Mean standardized 0.0002796RMS standardized 0.06106

Figure 5.Cross-validation scatter plot o the surace data.


8/14

581


100

200

300

400

100

200

300

400

35.519100

35.518600

35.518200

4402000

S

N

E

W

S-11

S-07

S-06S-05

S-22

S-20

S-23S-09

S-08S-17

S-14 S-26

S-13

S-16S-12S-10S-15S-03

S-04S-19

S-21

S-18S-25S-24

S-27

S-01S-02

S-11

S-07 S-06 S-05 S-22

S-20

S-23S-09

S-08S-17

S-14 S-26S-13

S-16S-12S-10S-15

S-03

S-04 S-19

S-21

S-18S-25

S-24

S-27

S-01S-02

Lithology

FAULT ZONE

MARBLE

METADIABASE

METASANDSTONE

ORE ZONE

SCHIST

SERPENTINITE

BB ADIT

YG ADIT

HDYU ADIT

HDK ADIT

a)

b)

Figure 6.(a) 3D topographic surace modeling with borehole points, adits, and the geological map o the Handeresi area (to avoidconusion, 500 m offset is applied to the geological map (Figure 2) and +500 m offset is applied to the 3D topographic suracemodeling along the z-axis). (b) 3D boreholes and adits o the Handeresi area.


9/14

582


(Pb%, Zn%, and Cu%) above the cut-off grades is moresuitable or choosing a target area. I there are not enoughboreholes without regular intervals, and i we do not detectthe ore zones more precisely using these data separately,intersecting the areas above the cut-off grades gives themost promising fields. Te probability o the presenceo ores in these fields is the greatest. Te purpose here isprimarily to evolve model files in which the values abovethe cut-off grade get 1 and the values below the cut-off

grade get 0, and then to determine the 1 value in thefile resulting rom multiplying these model files with eachother. In this latest model file, the areas having a 1 valueindicate intersection o above the cut-off grade o Pb%,Zn%, and Cu%. In order to determine intersecting area(s)with the help o some arithmetic operations, new modelsneed to be established. Tese operations are describedbelow.

Rockworks Utilities includes several modeling tools,such as generating or making changes to a solid model.Tese are displayed under the Solid menu. Te Solid/Boolean Operations/Boolean Conversiontool converts the

real number solid model file to a Boolean (true/alse)-type solid model file. In this process, the tool assigns a1 i G-values o nodes all within a user-defined rangeor assigns a 0 i they do not (RockWare 2006). In thisstudy, the percentage value o the elements is assigned to

each solid model file as a G-value. Te Pb% model file(Pb.mod) is chosen in the Boolean Conversiontool, and a

value o 1 is assigned to 7% (Pb cut-off grade) and highervalues.All values below 7% are accepted as 0 and a newmodel file consisting o Boolean values (Pb_boolean.mod)is created.All o these processes are applied separately toZn% values with a 4% cut-off grade and Cu% values witha 0.3% cut-off grade, and Boolean model files are created(Zn_boolean.mod and Cu_boolean.mod, respectively).

For the next step, the Solid/Mathtool, which includesthe arithmetic operation, is applied to solid models.Te

? ?

LithologyFAULT ZONE

MARBLE

METADIABASE

METASANDSTONE

ORE ZONESCHIST

SERPENTINITE

Figure 7.opographic surace modeling, boreholes, and ORE ZONE, which is modeledwith solid modeling o the Handeresi area (to avoid conusion, +500 m offset is applied

to boreholes and upper surace modeling image along to z- axis); side o view: rom NE.

able 2. Te parameters o the IDW interpolation method.

Weighting exponent 4

Max. points per voxel 64

Max. points per borehole 32

Sector width 90Sector height 90

Pb%

Zn%

Cu%

Figure 8.3D grade model o the Pb%, Zn%, and Cu% distributionsin the subsurace environment; side o view is the same in Figure9. See Figure 11 or colored interval legends.


10/14

583


options (Model&Model,Model&Constant,and Resample)within the Solid/Math tool are applied to arithmeticoperations on the values in the solid model files previouslycreated; this generates a new solid file (RockWare 2006).

Te Model&Model tool applies arithmetic operations tothe values o 2 model files and creates a new model file.In this study, the multiplication operation is applied toPb_boolean.mod with Zn_boolean.mod files.As a resulto the multiplication operation, the areas that include

1 values in both files (Pb and Zn cut-off grades and thehigher values) indicate 1 values, while the other areasindicate 0 values in the newly created Boolean modelfile (Pb_Zn_boolean.mod). Te multiplication operationis then applied or the generated file (Pb_Zn_boolean.mod) and the Cu_boolean.mod file.Te created file (Pb_Zn_Cu_boolean.mod) at the end o this process includes1 values that are assigned to Pb%, Zn%, and Cu%cut-off grades and higher values; the others include 0

values.Te visualization o this modeling and schematicrepresentations o these processes are shown in Figures 9band 10, respectively.

In order to observe spatial distribution modeling in 2D,2 cross-section lines (A-A and B-B) are drawn containingthe boreholes that take place in the possible ore zone areas(see Figure 2 or cross-section lines). In the first cross-section (A-A), distribution o the ore zones is observedbetween and around S-01 and S-02 boreholes at elevationso about 225300 m and around the S-17 borehole atelevations o 350375 m. In the second cross-section (B-B), distribution o the ore zones is observed around theS-23 borehole at elevations o about 215240 m (Figures11a and 11b).HDK and HDYU adits are situated near theS-01 and S-02 boreholes and the YG adit appears between

the S-18 and S-01 boreholes in cross-section A-A (Figure6). Tere are not any operating ore zone(s) and adits in theore zones near the S-17 borehole in cross-section A-A ornear the S-23 borehole in cross-section B-B.

In the modeling study, it is important to correlate withmeasured values and predicted values, which are revealed

Pb.mod

Zn.mod

Cu.mod

a)

b)

Figure 9.(a) Te view o the Pb.mod, Zn.mod, and Cu.modfiles afer the solid modeling process is applied. (b) Visualizationo the intersected zone afer the Boolean-type solid modeling andmathematical operations are applied to Pb.mod, Zn.mod, and

Cu.mod files.

x =

x =

0 1 1 0 0

0 1 1 1 0

0 1 1 1 1

1 1 1 0 0

0 0 0 0 0

0 0 1 1 0

0 1 1 1 0

0 1 1 1 1

0 1 1 1 1

0 0 1 1 1

0 0 1 0 0

0 1 1 1 0

0 1 1 1 1

0 1 1 0 0

0 0 0 0 0

0 0 1 0 0

0 1 1 1 0

0 1 1 1 1

0 1 1 0 0

0 0 0 0 0

0 1 1 1 0

0 1 1 0 0

0 1 1 0 0

0 0 1 1 0

0 1 1 1 0

0 0 1 0 0

0 1 1 0 0

0 1 1 0 0

0 0 1 0 0

0 0 0 0 0

Pb_boolean.mod

Pb_Zn_boolean.mod Cu_boolean.mod Pb_Zn_Cu_boolean.mod

Zn_boolean.mod Pb_Zn_boolean.mod

Figure 10. Schematic representations o the mathematical operations that areimplemented to Boolean-type files. Te 1 values indicate levels above the cut-offgrades, while 0 values indicate levels below them (the values on these figures werearbitrarily selected).


11/14

584


with the IDW interpolation method. Te grid files arecreated using the arbitrarily selected 275-m elevation data(x, y coordinate values and g values) rom the model filesand then the created grid files are cross-validated usingthe Geostatistical Analyst ool in ArcGis9 commercialsofware. Either the R2values are very close to the 1 valueor the RMS values in the prediction errors get relatively low

values, which indicates the accuracy o the interpolationmethod (Figure 12; able 3).

6. ConclusionsWith the help o computer technology, 3D subsuracemodeling has been used quite requently in the field omineral exploration in recent years. Te large amounto data obtained at regular spaces plays an importantrole in detecting the subsurace structures. In addition,reviewing previously collected insufficient data is veryimportant in order to both clariy previously missedstructures and ocus on the work underway in specific

areas. Furthermore, dense vegetation fields and/or high-slope areas make prospecting and geophysical studies

difficult. In these areas, to determine the potential areas onthe surace and/or at the depths, the modeling study usingborehole data is needed. Tereore, in this study, a targetarea that has irregularly spaced and insufficient boreholes,besides being explored actively at the present time and alsocovered with intensive vegetation, has been chosen. Teuse o active mining works is very important to compare

the results.As a result o the geostatistical analysis o the topographicdata creating DEMs, the appropriate variogram type isdetermined and the results are checked mutually usingthe cross-validation procedures. Consequently, a DEM iscreated with the kriging interpolation method using theExponential with nugget variogram.

In the Handeresi area, there are 27 boreholes thatare intensively observed in 2 different areas; there are noother boreholes outside o these areas. For this reason,the complete subsurace modeling, using all lithologic

variables, could not be implemented. Instead, the ore

zones and Pb%, Zn%, and Cu% analytical results weremodeled together. In addition, the aults in the area

A A'

0.0

100.0

S-18 0.0S-010.0

S-02

0.0

100.0

S-170.0

100.0

S-08

100.0

200.0

300.0

400.0

100.0

200.0

300.0

400.0

0.0 100.0 200.0 300.0 400.0

Cross-Secton (Pb%) A-A'B B'

0.0

100.0

S-27

0.0

100.0

200.0

S-23

0.0

100.0

S-06

100.0

200.0

300.0

400.0

100.0

200.0

300.0

400.0

0.0 100.0 200.0 300.0

Cross-Secton (Pb%) B-B'

22.020.018.016.014.012.010.08.06.04.02.00.0A A'

0.0

100.0

S-18 0.0S-010.0S-02

0.0

100.0

S-170.0

100.0

S-08

100.0

200.0

300.0

400.0

100.0

200.0

300.0

400.0

0.0 100.0 200.0 300.0 400.0

Cross-Secton (Zn%) A-A'B B'

0.0

100.0

S-27

0.0

100.0

200.0

S-23

0.0

100.0

S-06

100.0

200.0

300.0

400.0

100.0

200.0

300.0

400.0

0.0 100.0 200.0 300.0

Cross-Secton (Zn%) B-B'

8.58.07.57.06.56.05.55.04.54.03.53.02.52.0

1.51.00.50.0

A A'

0.0

100.0

S-18 0.0S-010.0S-02

0.0

100.0

S-170.0

100.0

S-08

100.0

200.0

300.0

400.0

100.0

200.0

300.0

400.0

0.0 100.0 200.0 300.0 400.0

Cross-Secton (Cu%) A-A'

B B'

0.0

100.0

S-27

0.0

100.0

200.0

S-23

0.0

100.0

100.

0

200.

0

300.0

400.0

100.

0

200.

0

300.0

400.0

0.0 100.0 200.0 300.0

Cross-Secton (Cu%) B-B'

2.42.22.01.81.61.41.21.00.80.60.40.20.0

Figure 11.Te cross-sections (a) (A-A) and (b) (B-B) o the Pb%, Zn%, and Cu% distributions at the subsurace (the red linesrepresent the topography). See Figure 2 or cross-section lines.


12/14

585


are very important with respect to controlling the orezones. Because the study area is covered with intensive

vegetation, the direction o aults on the surace is detectedwith aerial photos (1/35,000 in scale) and a geologic mapo the area drawn by Ycelay (1976). Because o manyaults and the dividing o the study area into small blocks,the subsurace modeling and controlling the continuityo the mineralizations become complicated. Moreover,in the borehole determinations o Ycelay (1976), theexact measurements related to the ault zones could notbe detected. Tat is the reason why the structural actorscould not be reflected in the modeling study.

Some nearly accurate results have been obtained usingthe IDWinterpolation method with the help o subsurace

solid modeling, but due to inadequacies in the data, notenough detailed inormation could be obtained. As aresult, there were 3 ore zones that were seen intensivelyin this area. One o them indicates the area in which theadits (Figure 6) are operating now (Zone 1, between andaround S-01 and S-02 boreholes, Figure 11a). Te other2 represent the area that is still not in operation (Zone 2,S-17 borehole, Figure 11a; Zone 3, S-23 borehole, Figure11b). In addition, the mineralization o Zone 1 is seen inthe areas situated between 225 and 300 m elevations andnear the HDK, HDYU, and YG adits. Te mineralizationso Zone 2 and Zone 3 are observed at elevations o about350375 m and 215240 m, respectively. In nearly o thesezones, there are not any adits and/or mining operations.Tese zones could be new potential areas.

Te zonal appearance o the ore mineralizationsinstead o along the specific zones is possible due tointensive aultings (Figures 11a and 11b). As mentioned

above, the structural actors could not be reflected in thesofware program because o data deficiencies. However,due to the act that the results are compared with the studyareas data and the operated ore zones in the area are alsodetected in this study, the applicability o the modeling isbrought out. In the uture, during surace and subsuraceexploration (e.g., boreholes, geophysical methods, welllogs) in this area, using these data and ocusing the studieson these zones will be much more economical with respectto the time and cost o mining operations. In addition, thedata previously collected but lacking enough inormationto be evaluated can be used or determining potential new

areas and/or developing areas being actively mined.

AcknowledgmentsTis work is a part o the PhD thesis o the first authorat the Geological Department o Ankara University,supervised by the second author, and was also supportedby the Scientific Research Office o Ankara University(BAPRO), Project No. 06B4343007. Te authors wish tothank the 2 anonymous reviewers or their helpul reviewsand also would like to thank Oreks Co. Ltd. or allowingaccess to the study area; Eli Akska, Alper Grbz, andKemal Mert nal or sharing their important views; andElaine Ambrose or revising the English in an early versiono the manuscript.

R = 0.9902

0.00

5.00

10.00

15.00

20.00

25.00

0.00 5.00 10.00 15.00 20.00

Pre

d

cte

dva

lues

Measured values

Pb% (IDW method)

R = 0.9901

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00

Pre

dcte

dva

lues

Measured values

Zn% (IDW method)

R = 0.9989

0.00

0.05

0.10

0.15

0.20

0.25

0.300.35

0.40

0.45

0.50

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50

Pre

dctvalues

Measured values

Cu% (IDW method)

Figure 12. Cross-validation scatter plot o the geochemicalanalyses data.

able 3. Te summary statistics o the prediction errors usingIDWinterpolation methods.

Prediction errors

Pb% Zn% Cu%

Samples 30,351 30,351 30,351

Mean 0.0000089 -0.0000166 0.0000011

RMS 0.3061 0.1581 0.003925


13/14

586


References

Akska, S. 2010. Yenice (anakkale) Blgesindeki Cu-Pb-Zn Oluumlar[Cu-Pb-Zn Occurrences o the Yenice (anakkale) Area]. PhDthesis, Ankara University, Ankara, urkey [unpublished].

Akska, S., Demirela, G., Sayl, .S. & Kucu, . 2010a. Fluid inclusionand S isotope systematics o some carbonate-related Pb-Zn-Cu mineralizations in NW Anatolia, urkey. In: Melos, V.,Marchev, P., Lakova, I. & Chatzipetros, A. (eds), GeologicaBalcanica Abstract Volume, p. 21.

Akska, S., Sayl, .S. & Demirela, G. 2010b. Te subsurace 3Dmodeling o the Handeresi (Kalkim-Canakkale) area, NWo urkey, Pb-Zn-Cu ore zones. EGU2010 General Assembly,Vienna, Abstracts, 1636.

Armstrong, M. 1998. Basic Linear Geostatistics.Springer, Berlin.

Binh, .Q. & Tuy, N.. 2008. Assessment o the influence ointerpolation techniques on the accuracy o digital elevation

model. VNU Journal o Science, Earth Sciences24, 176183.

Bistacchi, A., Massironi, M., Dal Piaz, G.V., Dal Piaz, G., Monopoli, B.,Schiavo, A. & offolon, G. 2008. 3D old and ault reconstructionwith uncertainty model: an example rom an Alpine tunnel casestudy. Computers & Geosciences34, 351372.

Brooker, P.I. 1986. A parametric study o robustness o kriging varianceas a unction o range and relative nugget effect or a sphericalsemivariogram.Mathematical Geology18, 477488.

Burrough, P.A. & McDonnell, R.A. 1998. Principles o GeographicalInormation Systems. Oxord University Press, Oxord.

aatay, A. 1980. Bat Anadolu kurun-inko yataklarnn jeoloji-

mineraloji etd ve kkenleri hakknda grler [Geology andmineralogy o western Anatolian lead-zinc deposits and somecomments about their genesis]. rkiye Jeoloji Kurumu Blteni23, 119132 [in urkish with English abstract].

etinkaya, N., Karul, B., nal, R. & Yenign, K. 1983. GeologicalReport on the anakkale-Yenice-Kalkm Handeresi Pb-Zn-Cu

Mineralization. General Directorate o Mineral Research andExploration (MA), Report No. 7822 [in urkish, unpublished].

Chaouai, N.E. & Fytas, K. 1991. A sensitivity analysis o search distanceand number o samples in indicator kriging. CIM Bulletin84,3743.

Chaplot, V., Darboux, F., Bourennane, H., Leguedois, S., Silvera, N. &Phachomphon, K. 2006. Accuracy o interpolation techniquesor the derivation o digital elevation models in relation tolandorm types and data density. Geomorphology77, 126141.

Cressie, N.A.C. 1990. Te origins o kriging.Mathematical Geology22,239252.

de Kemp, E.A. 2000. 3-D visualization o structural field data: examplesrom the Archean Caopatina Formation, Abitibi greenstone belt,Qubec, Canada. Computers & Geosciences26, 509530.

Demirela, G., Akiska, S., Sayili, .S. & Kucu, . 2010. Silicate andsulfide mineral chemistry o some carbonate related Pb-Zn-Cumineralizations and their effects on ore genesis in NW Anatolia,urkey. In:Melos, V., Marchev, P., Lakova, I. & Chatzipetros,A. (eds), Geologica Balcanica Abstract Volume, p. 91.

Elkadi, A.S. & Huisman, M. 2002. 3D-GSIS geotechnical modelingo tunnel intersection in sof ground: the second Heinenoordtunnel, Netherlands. unnelling and Underground Spaceechnology17, 363369.

Feltrin, L., McLellan, J.G. & Oliver, N.H.S. 2009. Modelling the giant,Zn-Pb-Ag century deposit, Queensland, Australia. Computers &Geosciences35, 108133.

Fenck, R. & Vajsblov, M. 2006. Parameters o interpolation methodso creation o digital model o landscape.Te 9thAGILE Conerenceon Geographic Inormation Science, Visegrad, Hungary, 374381.

Galera, C., ennis, C., Moretti, I. & Mallet, J.L. 2003. Constructiono coherent 3D geological blocks. Computers & Geosciences29,971984.

Hack, R., Orlic, B., zmutlu, S., Zhu, S. & Rengers, N. 2006. Treeand more dimensional modelling in geo-engineering. Bulletin o

Engineering Geology and the Environment65

, 143153.Harris, N.B.W., Kelly, S. & Okay, A.. 1994. Post collision magmatism

and tectonics in northwest Anatolia. Contributions to Mineralogyand Petrology117, 214252.

Houlding, S.W. 1992. Te application o new 3D computer modelingtechniques to mining. In:urner, A.K. (ed), Tree-Dimensional

Modelling with Geoscientific Inormation Systems. KluwerAcademic Publishers, Dordrecht, the Netherlands, 303326.

Houlding, S.W. 1994. 3D Geoscience Modeling: Computer echniques orGeological Characterization. Springer, Berlin.

Johnston, K., Ver Hoe, J.M., Krivoruchko, K. & Lucas, N. 2001. UsingArcGIS Geostatistical Analyst.ESRI Press, Redlands, CA, USA.

Kaumann, O. & Martin, . 2008. 3D geological modelling romboreholes, cross-sections and geological maps, applicationover ormer natural gas storages in coal mines. Computers &Geosciences34, 278290.

Krige, D.G. 1951. A statistical approach to some basic mine valuationproblems on the Witwatersrand. Journal o the Chemical,

Metallurgical and Mining Society o South Arica52, 119139.

Krushensky, R.D. 1976. Neogene calc-alkaline extrusive and intrusiverocks o the Karalar-Yeiller area, Northwest Anatolia. BulletinVolcanologique39, 336360.

Lam, N.S. 1983. Spatial interpolation methods review. Te American

Cartographer10, 129149.Li, J. & Heap, A. 2008. A Review o Spatial Interpolation Methods or

Environmental Scientists. Record 2008/23, Geoscience Australia,Canberra.

Matheron, G. 1960. Krigeage dun Panneau Rectangulaire par saPriphrie. Note Geostatistique No. 28, CG. Ecole des Mines deParis, Paris.

Ming, J., Pan, M., Qu, H. & Ge, Z. 2010. GSIS: A 3D geological multi-body modeling system rom netty cross-sections with topology.Computers & Geosciences36, 756767.

Okay, A.., Siyako, M. & Brkan, K.A. 1990. Biga Yarmadasnnjeolojisi ve tektonik evrimi [Geology and tectonic evolution othe Biga Peninsula]. rkiye Petrol Jeologlar Dernei Blteni2,83121 [in urkish with English abstract].


14/14

587


Okay, A.., Satr, M., Maluski, H., Siyako, M., Monie, P., Metzger, R.& Akyz S. 1996. Paleo- and Neo-ethyan events in northwesturkey: geological and geochronological constraints. In:Yin, A.& Harrison, M. (eds), ectonics o Asia. Cambridge UniversityPress, Cambridge, 420441.

Okay, A.., Satr, M. & Siebel, W. 2006. Pre-Alpide orogenic events inthe Eastern Mediterranean region. In:Gee, D.G. & Stephenson,R.A. (eds), European Lithosphere Dynamics. Geological SocietyLondon, Memoirs 32, 389405.

zmutlu, S. & Hack, H.R.G.K. 1998. Excavatability evaluation andclassification with knowledge based GIS. In:Proceedings o 8thInternational IAEG Congress, Vancouver, 591598.

zmutlu, S. & Hack, R. 2003. 3D modelling system or groundengineering. In:Rosenbaum, M.S. & urner, A.K. (eds), NewParadigms in Subsurace Prediction Characterization o the

Shallow Subsurace Implications or Urban Inrastructure and

Environmental Assessment, Lecture Notes in Earth Sciences, Vol.

99, Springer, Berlin, 253260.Peralvo, M. 2004. Influence o DEM interpolation methods in drainage

analysis. GIS Hydro 04, University o exas, Austin, exas.

Rahman, K.M. 2007. Settlement Prediction on the Basis o a 3DSubsurace Model, Case Study: Reeuwijk Area, the Netherlands.MSc thesis, International Institute or Geo-InormationScience and Earth Observation, Enschede, the Netherlands[unpublished].

Renard, P. & Courrioux, G. 1994. Tree-dimensional geometricmodeling o a aulted domain: the Soultz Horst example (Alsace,France). Computer & Geosciences20, 13791390.

Rengers, N., Hack, R., Huisman, M., Slob S. & Zigterman, W. 2002.Inormation technology applied to engineering geology. In:van Rooy, J.L. & Jermy, C.A. (eds), Engineering Geology orDeveloping Countries Proceedings o 9th Congress o the

International Association or Engineering Geology and the

Environment, Durban, South Arica, 121143.

RockWare. 2006. RockWorks V.2006 - Earth Science and GIS Sofware.RockWare, Golden, CO, USA. Available at http://www.rockware.com/assets/products/165/downloads/documentation/11/rw2006_manual.pd [accessed 16 October 2011].

Sara, C. & ercan, A.E. 1996. Grade and reserve estimation o theulovasi borate deposit by block kriging. International Geology

Review38, 832837.

Sims, D.L. 1992. Application o 3D geoscientific modelling orhydrocarbon exploration. In: urner, A.K. (ed), Tree-Dimensional Modelling with Geoscientific Inormation Systems.Kluwer Academic Publishers, Dordrecht, the Netherlands,285290.

Sweetkind, D.S., aylor, E.M., McCabe, C.A., Langenhem, V.E. &McLaughlin, R.J. 2010. Tree-dimensional geologic modeling othe Santa Rosa Plain, Caliornia. Geosphere 6, 237274.

ercan, A.E. 1996. Maden yataklar snr belirsizliinin indikatrkriging ile deerlendirilmesi ve Sivas-Kangal-Kalburayrkmr yatanda bir uygulama [Assessment o boundaryuncertainty o ore deposits by indicator kriging and itsapplication to the coal deposit o Kalburayr, Kangal, Sivas].

Madencilik35, 312 [in urkish with English abstract].

ercan, A.E. & Sara, C. 1998.Maden yataklarnn deerlendirilmesindejeoistatistiksel yntemler [Geostatistical methods in the

evaluation of ore deposits]. MMOB Jeoloji MhendisleriOdas Yaynlar 48, urkey [in urkish].

uan, A.E. 1993. Karaydn Ky (Yenice-anakkale) evresininjeolojik ve petrografik zellikleri ile kurun-inko zuhurlarnn

jenetik incelemesi [Geological and Petrographical Characteristics

of the Karaaydn Region (Yenice-anakkale) with the Genetic

Investigations of Lead-Zinc]. PhD thesis, Seluk University,urkey [in urkish, unpublished].

urner, A.K. 1992. Tree-Dimensional Modelling with Geoscientific

Information Systems.Kluwer Academic Publishers, Dordrecht,the Netherlands.

Veldkamp, J.G., Hack, H.R.G.K., zmutlu, S., Hendriks, M.A.N.,Kronieger, R. & Van Deen, J.K. 2001. Combination o 3D-GISand FEM modelling o the 2nd Heinenoord unnel, theNetherlands. In: Proceedings of the International Symposium onEngineering Geological Problems of Urban Areas (EngGeolCity),Ekaterinburg, Russia, 18.

Wang, G., Zhang, S., Yan, C., Song, Y., Sun, Y., Li, D. & Xu, F. 2011.Mineral potential targeting and resource assessment basedon 3D geological modeling in Luanchuan region, China.Computers & Geosciences37, 19761988.

Xue, Y., Sun, M. & Ma, A. 2004. On the reconstruction o three-dimensional complex geological objects using Delaunaytriangulation. Future Generation Computer Systems20, 12271234.

Ycelay, M.A. 1976. Te Etude of anakkale-Kalkm-Handeresi Pb-Zn-Cu Area. General Directorate o Mineral Research andExploration (MA), Report No. 5720 [in urkish, unpublished].

Zanchi, A., Francesca, S., Steano, Z., Simone, S. & Graziano, G. 2009.3D reconstruction o complex geological bodies: examplesrom the Alps. Computers & Geosciences35, 4969.

Zhu, S., Hack, R., urner, A.K. & Hale, M. 2003. How ar willuncertainty o the subsurace limit the sustainability planning othe subsurace? In: Proceedings of the Sustainable Development& Management of the Subsurface (SDMS) Conference, Utrecht,the Netherlands, 203210.

Zimmerman, D., Pavlik, C., Ruggles, A. & Armstrong, P. 1999. Anexperimental comparison o ordinary and universal krigingand inverse distance weighting. Mathematical Geology 31,375390.

Documents

62 Three-dimensional Subsurface Modeling of Mineralization a Case Study From the Handeresi (Çanakkale, NW Turkey) Pb-Zn-Cu Deposit