ABSTRACTParion is one of the most important se lements located in the ancient Troas region, in

which the city of Troy was the center. Many remarkable and precious archaeologicalremains have been unearthed so far which point out the city’s importance during theHellenistic and Roman Age. In this study, a first a empt to obtain high resolution imagesof the subsurface of Parion to guide the archaeological trenches was made by an initialgeophysical survey applying Electrical Resistivity Tomography (ERT) technique. Theapparent resistivity data, collected using pole‑dipole electrode configuration along 11transects, were inverted by two‑ and three‑dimensional smoothness‑constrained leastsquares algorithms. Relatively compatible results were obtained from two inversionprocesses. Parallel transects showed the resistivity distribution in three‑dimensionalimages and thus both the horizontal and vertical extents of the anomalous zones weredisplayed. Additionally, some high anomaly zones located at the end of the first sixtransects were backed up by archaeological trenches. Thus, taking into account thesefindings, the other resistivity anomalies located at the different parts of the surveyed areaare thought to be the most promising locations for archaeological excavations.

GEOPHYSICAL IMAGING SURVEY IN THE SOUTHNECROPOLIS AT THE ANCIENT CITY OF PARION

(KEMER ‑ BIGA), NORTHWESTERN ANATOLIA,TURKEY: PRELIMINARY RESULTS

Yunus Levent Ekinci1, Mehmet Ali Kaya2, Cevat Başaran3, Hasan Kasapoğlu3, Alper Demirci1 and Cemal Durgut4

1 Çanakkale Onsekiz Mart University, Department of Geophysical Engineering, TR‑17020, Çanakkale‑Turkey

2 Okul sk. No: 11, TR‑42960, Taşkent‑Konya‑Turkey3 Atatürk University, Department of Archaeology, TR‑25240, Erzurum‑Turkey

4 JeoTomografi Engineering Company, TR‑09020, Aydın‑Turkey

Corresponding author: ylekinci@comu.edu.trReceived: 15/3/2012Accepted: 15/4/2012

KEYWORDS: Geophysics, archaeology, resistivity, inversion, necropolis, volumetric imaging, Parion

Mediterranean Arhaeology and Archaeometry, Vol. 12, No 2, pp.145‑157Copyright @ 2012 MAA

Printed in Greece. All rights reserved.

INTRODUCTION

Archaeological studies involveexcavations which are time consuming andthese excavations also require a hugeexpenditure of physical effort.Unfortunately, this effort may not be verycost‑effective due to the risks of damagingor missing the burial archaeologicalremains in some cases. On the other hand,archaeological features have some certainphysical properties (i.e. magnetic suscep‑tibility, electrical resistivity andconductivity, dielectric constant) which canbe measured from the earth surface. Thususeful information about some physicalparameters of the buried archaeologicalstructures such as location, depth, size,thickness, position and extent can beobtained prior to excavations by using non‑invasive and non‑destructive geophysicalimaging techniques.

The first geophysical survey in an archae‑ological site was carried out as early as thelate 1940s (Atkinson, 1952). However onlysince the 1980s the geophysical methodshave been widely performed on manyarchaeological sites. It is known thesuccessful use of geophysical prospection inarchaeological sites to unearth buriedantiquities (e.g. Hesse et al., 1986; Scollar etal., 1986; Griffiths and Barker, 1994; Tsokaset al., 1994; Neubauer, 1997; Gaffney et al.,2000; De Domenico et al., 2006; Leucci et al.,2007; Drahor et al., 2008). The mostcommonly applied geophysical methodsused for archaeological explorations areelectrical resistivity, magnetic, groundpenetrating radar and electromagneticconductivity. These subsurface imagingtechniques are carried out easily andquickly on the ground surface withoutdisturbing or damaging the buried archae‑ological features. Additionally, theadvantages mentioned above increase thepopularity of non‑destructive geophysicalsurveys at archaeological sites day by day(e.g. Savvaidis et al., 1999; Diamanti et al.,

2005; Papadopoulos et al., 2006; Ekinci andKaya, 2007; Büyüksaraç et al., 2008). Directcurrent resistivity method is one of the mostcommonly applied geophysical imagingtechniques for shallow investigations.Electrical Resistivity Tomography (ERT),the modified version of conventionalvertical electrical sounding, is based onreconstruction modeling techniques and isused to obtain high resolution resistivityimages of the subsurface. The aim of thetechnique is to collect apparent resistivitydata from the subsurface along a transectusing a predefined electrode array in orderto generate a model section images. ERT hasbecome an increasingly efficient tool toinvestigate buried archaeological featuresdue to the ability of detecting walls, voids,graves and some other man‑madestructures. The low cost of the investigationand the possible resistivity contrast betweenthe archaeological structures and thesurrounding soil are also the mainadvantages of the method (Ekinci and Kaya,2007). Additionally, ERT has gained moreacceptances in recent years due to thetechnological advances in resistivity‑metersystems and the two‑ and three‑dimensional forward and inversionmodeling algorithms. Apparent resistivitydata are measured along a predeterminedline by sequence of a selected electrodeconfiguration for building up a two‑dimensional pseudo‑section or obtainedfrom a selected area by differentarrangements of electrode configurationsfor three‑dimensional surveys (Loke, 2001;Papadopoulos et al., 2006; Drahor et al.,2008). A er the data acquisition process,apparent resistivity data are inverted bytwo‑ or three‑dimensional optimizationalgorithms in order to obtain true resistivitydistribution of the subsurface. At the laststage, volumetric representations,orthogonal and depth slices are generatedto display the anomalous zones in thegeoelectrical model sections.

The presented work was carried out to

146 YUNUS LEVENT EKINCI et al

explore and locate the buried man‑madearchaeological structures at the southnecropolis area of Parion ancient city (Biga‑Turkey). It was thought that the results ofthe geoelectrical imaging survey wouldguide to the excavations and give insightinto the pa erning of unexcavated parts ofthe necropolis. Measured apparentresistivity data gathered along each transectwere inverted by two‑dimensional inversionalgorithm and the results were illustratedjointly (quasi‑three‑dimensional approach).Additionally, by combining parallel two‑dimensional apparent resistivity data setswe used a three‑dimensional inversionalgorithm (semi‑fully three‑dimensionalapproach). More accurate interpretation wasaimed by using the results of both twoalgorithms ( Papadopoulos et al., 2006; Berge and Drahor, 2011). In order to carry out a more detailed analysis and improve the interpretation,

were produced by using two‑dimensional orthogonal and depth slices. These images were generated based on multi‑dimensional gridding processes forvolumetric data. The technique allowsdisplaying of orthogonal planes in thedirection at predetermined positions.Finally, resistivity tomograms displayedmany high anomaly zones and theanomalies located at the end of the six lineswere backed up by archaeological trenches.Due to this evidence, the other resistivityanomalies in the tomograms are thought tobe the most promising locations for thesubsequent seasons’ excavations.

SITE DESCRIPTION AND HISTORICAL BACKGROUND

Parion is located about 90 km to the city ofÇanakkale, NW Turkey (Fig. 1a). The villageof Kemer (Fig. 1b) which belongs to theadministrative district of Biga, partially

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION 147

Figure 1. Location map of the ancient city of Parion (a) and the view of the village Kemer (b).

148 YUNUS LEVENT EKINCI et al

Figure 2. Map of the Ancient Troas Region.

Figure 3. The locations of the focused archaeological studies in the city of Parion.

archaeology museum of Çanakkale in 2004,and has been under systematic excavationsince 2005. Many graves and remains ofstructures, which are thought to belong tothe era from 6th century BC to the end of theRoman period, have been brought to light byarchaeological excavations in southnecropolis so far (Fig. 4).

covers the city. Parion is surrounded byabout 8 km city walls. According to Başaran(2008) the city is considered as a city ofAncient Troas Region (Fig. 2). Theatre,Roman Bath, Slope House, Stone Tower andNecropolis are the focused localities inParion (Fig. 3). The city was first investigatedduring rescue excavations, completed by the

There are several opinions about thename of Parion. One idea suggests that thecity was named a er Parion, son of Iasonand Demetria, emigrants from Erythraili,who established the city (Leaf, 1923;Bonacasa, 1976; Frish, 1983; Başaran, 1998).Another view holds that the name wasoriginated from the people emigrated fromParos in Colonization Era (Başaran, 2001;Avram, 2004). The third view claims that thecity was named a er Paris, the youngest sonof the King Priam, and therefore it isthought that Parion means “the city of Paris”(Başaran, 1998 and 2005). Although theinformation about the early period of thecity is limited, the historical data of Parionbelong to the 6th century BC. It is known thatthe city came under the rule of Persiandomination a er the end of the LydianKingdom in 546 BC (Mansel, 1988; Başaran,2002). Later on, Parion became a member ofAtika‑Delos Sea Union which was foundedin 479 ‑ 478 BC to emancipate the citiesunder Persian rule (Mansel, 1998; Başaran,

2002). During the Peloponnesus Wars (431 ‑404 BC) the people of Parion and Athensfought against the Spartans. Following theinvasion of Galatians between 278 ‑ 277 BC,the city was captured by PergamonKingdom in 271 BC (Mansel, 1988; Başaran,2002). A er Pergamon had come under therule of Roman in 133 BC, Parion joined toRoman Empire as a political authority andduring the first Mithridates wars the citygained a half‑independent political status(Başaran, 2002). Additionally, Parion washighly valued in Roman Era and declared as“the favored city” with the title of “ColoniaPariana Iulia Augusta” by Augustus (27 BC‑ 14 AD) (Magie, 1950; Başaran, 2001 and2006).

DATA ACQUISITION AND PROCESSING

Geoelectrical imaging survey at southnecropolis was carried out using the pole–dipole electrode array configuration. As is

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION 149

Figure 4. Sketch plan of the necropolis showing some archaeological remains brought to light by theexcavations since 2005. Yellow grids (two middle squares at C) show the trenches coinciding with

the six geoelectrical resistivity transects.

known, the pole‑dipole array geometry hasrelatively good horizontal coverage, but it isconsidered to be more effective than thedipole‑dipole array because the movementof one transmi er electrode is sufficient andthe array produces considerably higherreceiver voltage (White et al., 2003). Themost obvious advantages of pole‑dipolearray over the dipole‑dipole and wenner‑schlumberger arrays are the speed of dataacquisition, superior depth of investigationand the volume of the data (White et al.,2003). Additionally, it is not as sensitive totelluric noise as the pole‑pole array (Loke,2001). Therefore pole‑dipole array wasconsidered to be the most suitable electrodearray for this survey. The apparentresistivity data were collected in an area of23 m by 20 m, over 11 parallel transects, each23 m long, with line spacing 2 m. There wereno significant topographical changes in thispart of the necropolis. Taking intoconsideration the archaeologicalinformation obtained from the trenches nextto the geophysical survey area at thenecropolis such as depth and dimension, thepenetration depth of the survey was set toabout 4.5 m with electrode spacing of 1 m.The remote current electrode was placedsufficiently far from the survey lines i.e morethan 66 m (more than 6 times the largest C1‑P1 distance used (n factor) on the surveyline). A total of 2057 apparent resistivityvalues were collected for 11 data levels (n=1to n=11) using the IRIS‑Syscal R1 Plusresistivity‑meter. In order to enhance thequality of the measured apparent resistivitydata, we set the number of stacks (repeatmeasurements) to 4 and it was increased to8 when the relative standard deviation of thestacked data was greater than 2 %. Theprocedure was applied to only a smallnumber of datum points since the relativestandard deviation values were less than 0.5% for a clear majority due to the low noiselevel of the surveyed site. Even though thepseudosection representation producedfrom apparent resistivity distribution may

give some information about the location ofthe causative sources, the parameters suchas size, depth, thickness and extent cannotbe estimated correctly. Thus we applied two‑and three‑dimensional inversion methods inorder to obtain true resistivity images of thesubsurface using Res2Dinv and Res3Dinvso ware products (Loke and Barker, 1996),respectively. The aim was to interpret moreaccurately by taking advantages of theresults of both two algorithms. Theseinversion algorithms, based on smoothness‑constrained least‑squares (deGroot‑Hedlinand Constable, 1990; Sasaki, 1992)implemented by a quasi‑Newtonoptimization technique (Loke and Barker,1996), iteratively calculate a resistivitymodel, trying to minimize the differencebetween the observed apparent resistivityvalues and those calculated from the model(Loke and Barker, 1996). The calculatedapparent resistivity values were obtained byfinite element method using 4 nodes perunit electrode spacing during the inversionprocess. The inverse model resistivitysections were selected at the iteration a erwhich the root mean square (RMS) error didnot change significantly. Satisfactory resultswere obtained a er 5 iterations for all linesand the RMS errors were between 0.8 ‑ 4.3%. The three‑dimensional inversion ofcollated resistivity data obtained from 11transects produced an inverse modelresistivity at the end of 5 iterations with aRMS error of 4.1 %. Due to low RMS errors,the obtained results can be considered as areliable representation of the true resistivitydistribution of subsurface. Finally, aMATLAB‑based includingvisualization functions was used to displaythe resistivity distribution of each profilejointly in three‑dimensional volumetricslices and depth sections. Thesevisualization techniques allowed displayingthe three‑dimensional volume in a range ofuser‑selected orthogonal slices of boundingsurfaces. Therefore the extents of theanomalous zones were easily displayed.

150 YUNUS LEVENT EKINCI et al

RESULTS AND DISCUSSION

The nearly WSW‑ENE trendingtomograms displayed a resistivitydistribution of a depth range of about 4.5 mand showed many remarkable highresistivity anomalies. The resistivitytomograms obtained from two‑ and three‑dimensional inversion processes areillustrated in Fig. 5a and b, respectively. Thejointly represented tomograms indicatedthat the overall resistivity range in thesurveyed area varies between about 20 and200 ohm‑m except a negligible amount ofdatum points. Test measurement carried

out on an area on the outer part of the sitewhere the existence of archaeologicalremains is not expected showed that thecovering material has a moderate resistivityvalues ranging between 25 and 84 ohm‑m.In the light of this information, highresistivity zones, i.e. more than 110 ohm‑m,were considered to be the possible traces ofthe man‑made archaeological burialstructures.

Some resistive zones were determined atthe end of the first six ERT slices (RZN 1through RZN 6) obtained from both twoinversion procedures (Fig. 5a and b). Thisanomaly zones showed that the results of

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION 151

Figure 5. Geoelectrical imaging of the surveyed area with orthogonal slices obtained from two‑dimensional inversion (a) and three‑dimensional inversion approach (b) (note that the long axes

were exaggerated for the illustrations).

two‑ and three‑ dimensional inversions arein accordance with each other about thelocation. However, it was observed thatthere is no certain agreement between the

two inversion procedures about the verticalextent of the anomalous zones. Thus, inorder to investigate these zones in a moredetailed manner, the inverted resistivity

data were demonstrated as depth slicesstarting at the elevation of 8.15 m down to4.15 m with an interval of 0.5 m (Fig. 6a andb). It can be clearly seen from Fig 6a that thehigh resistivity zone, marked by whitearrows, is located between the elevations ofabout 6.65 and 4.65 m. The decrease in theintensity of the anomaly from mentionedlevel to downward suggests that thepossible man‑made archaeologicalstructures do not exist below the elevationsof about 5.15 – 4.65 m. On contrast to two‑dimensional inversion results, depth slices(Fig. 6b) obtained from three‑dimensionalinversion procedure showed that theanomalous zone, highlighted by whitearrows, continues downward vertically toan unknown depth level without decreasingintensity (maybe more than 5 m below theground surface). Considering theinformation obtained from the results of the

excavations next to the investigation area,the results of the two‑dimensional inversionseemed to be more logical. These anomalyzones coincide with two grids of the archae‑ological trenches conducted by excavation(highlighted by yellow grids in Fig. 4). Somegraves and archaeological remains wereunearthed in these grids since 2009. Theview of the surveyed area before and a erthe excavations, trenches, and close‑upviews of the some unearthed material inthese grids are shown in Fig. 7a, b, c, d ande, respectively. These grids have a surfaceelevation of 8.65 m and 7 tile graves, datedas 1‑2th centuries AD, made of flat ortrapezoidal (convex) tiles were foundbetween the elevations of 8.65 ‑ 7.85 m(Kasapoğlu, 2007) (Fig.8a). The anomaly ofthese thin remains can be clearly seen fromFig. 9 highlighted with thick white arrow inthe resistivity tomogram of RZN 1.

152 YUNUS LEVENT EKINCI et al

Figure 6. Geoelectrical imaging of the surveyed area with depth slices obtained from two‑dimensional inversion (a) and three‑dimensional inversion approach (b) (note that the vertical axes

were exaggerated for the illustrations).

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION 153

Figure 8. Three‑dimensional illustrations of unearthed archaeological structures coincidingwith high resistivity zones (a, b and c).

Figure 7. The views of the surveyed area before and a er the excavations (a, b), archaeological trenches (c), and the close‑up views of the some unearthed materials coinciding with high

resistivity zones (d, e).

Additionally, 11 tile graves, mostly dated asthe 1st century AD, that had been formed byflat and trapezoidal (convex) tiles like thegraves nearly above them were found at theelevations of 7.85 – 6.65 m (Fig. 8b), andthese parts of the trenches represent slightlyhigh resistivity values in the tomogram ofRZN 1, marked with thin arrow (Fig. 9).Therewithal, at the elevations of 6.45 – 4.95m one stone‑chest tomb (Kasapoğlu, 2007)and two marble‑chest tombs, which arethought to belong to the Hellenistic Period,were uncovered. Some golden jewelry werealso found in these tombs. High resistivity

zones in RZN 2 and 3 highlighted witharrows point out these remains (Fig. 9). Atthe elevations of 5.95 – 4.95 m just in thesouth of those graves, an architecturalremain formed by sandstone plates wasunearthed. The traces of this platformmarked by arrows in Fig. 9 can be clearlyseen from the tomograms of RZN 3 throughRZN 6. The dimension of the structure isabout 5.50 x 3.30 x 1.00 m and was made offour rows of sandstone plates, each havingabout 0.25 m thickness, placed successively(Fig. 8c). Due to the knowledge obtainedfrom the graves unearthed around this

154 YUNUS LEVENT EKINCI et al

Figure 9. Zoomed views of the first sixgeoelectrical resistivity slices coinciding witharchaeological trenches (Note that the long axiswas exaggerated for the illustration).

platform, it is believed that this remainbelongs to the 6 – 5th century BC. Althoughthe meaning of this platform is yet notcertain, the information that will be obtainedfrom the next seasons’ excavations will helpus to understand the usage purpose of thisarchitectural structure.

CONCLUSIONS

This paper presents the preliminaryresults of the first archaeogeophysicalsurvey performed in the south necropolis atthe ancient city of Parion. The geophysicalstudy focused on quasi and semi‑fullythree‑dimensional evaluation ofgeoelectrical imaging data, composed ofrelatively dense parallel two‑dimensionaltransects in reconstructing the subsurfaceelectrical resistivity properties of thenecropolis area, by ERT technique. Bytaking the advantages of two‑ and three‑dimensional inversion approaches and alsothe multi‑dimensional volumetricvisualization techniques, the producedelectrical resistivity tomograms wereinterpreted in a more accurate manner.Additionally, excavated archaeologicaltrenches in two neighboring gridscoinciding with the geophysical survey arearevealed the high degree of accordancebetween the man‑made archaeologicalstructures and the geoelectrical anomaly

zones having high resistivity values. In thelight of the obtained results, the other highresistivity anomaly zones observed inresistivity tomograms marked by symbolsare thought to be the buried archaeologicalremains with a high degree of probability.Thus, these zones are the most promisinglocations for the following archaeologicalexcavations. Summarizing, it can beconcluded that the non‑destructive andnon‑intrusive ERT technique and the use ofmulti‑dimensional volumetric visualizationmethods are very suitable and effectivestrategy for the exploration of buriedarchaeological remains.

ACKNOWLEDGEMENTS

Thanks are due to Mrs. Rezzan Ekinci(geophysicist), Mr. Can Ertekin (geologist)and Mr. Taner Gürer (geophysicist) for theirvaluable helps in preparing the high qualityimages. We are grateful to Ms. CananAlbayrak (archaeologist) for her great effortduring the data acquisition stage of thegeoelectrical resistivity imaging survey.Excavation team is thanked for their helpsand hospitality. We are also indebted to twoanonymous referees for their constructivecomments. The location maps shown in Fig.1a and Fig. 2 were generated by usingGeneric Mapping Tools (GMT) (Wessel andSmith, 1995).

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION. 155

REFERENCES

Atkinson, R.J.C., (1952). Methodes electriques de prospection en arceologie, A. Lamining(Ed.), La Décoverte Du Passé. Picard, 59‑70.

Avram, A., (2004). The propontic coast of Asia Minor, An inventory of archaic and classical poleis.Oxford University Press.

Başaran, C., (1998). Parion 1997 araştırmaları. XVI. Araştırma Toplantısı Sonuçları, 349‑364(in Turkish).

Başaran, C., (2001). Parion’dan iki gladiator steli, Günışığında Anadolu. CevdetBayburtluoğlu Için Yazılar, 19‑24 (in Turkish).

Başaran, C., (2002). Geçmişten günümüze Bayramiç, tarihi – coğrafyası ve arkeolojisi. T.C.Kültür Bakanlığı Yayınları (in Turkish).

Başaran, C., (2005). Jewellery center in antique era: Parion. Goldnews, 153, 106‑111.

Başaran, C., (2006). Excavating Hellenistic Parion: A royal necropolis in Turkey. Minerva,17, 36‑37.

Başaran, C., (2008). Parion’dan Persia’ya Yol Gider, Euergetes. Prof.Dr. Haluk Abbasoğlu’na65.Yaş Armağanı, 1, 134 (in Turkish).

Berge, M.A., and Drahor, M.G., (2011). Electrical resistivity tomography investigations ofmultilayered archaeological se lements: Part II – A case from old SmyrnaHöyük,Turkey. Archaeological Prospection, 18, 291‑302.

Bonacasa, N., (1976). Princeton encyclopedia of classical sites. Princeton University Press.Büyüksaraç, A., Arısoy, M.Ö., Bektaş, Ö., Koçak, O., and Çay, T., (2008). Determination of

grave locations in Dedemezari Necropolis (Western Turkey) using magnetic fieldderivatives. Archaeological Prospection, 15, 267‑283.

De Domenico, D., Giannino, F., Leucci, G., and Bo ari, C., (2006). Integrated geophysicalsurveys at the archaeological site of Tindari (Sicily, Italy). Journal of ArchaeologicalScience, 33, 961‑970.

deGroot‑Hedlin, C., and Constable, S.C., (1990). Occam’s inversion to generate smooth,two‑dimensional models fom magnetotelluric data. Geophysics, 55, 1613‑1624.

Diamanti, N.G., Tsokas, G.N., Tsourlos, P.I., and Vafidis, A., (2005). Integratedinterpretation of geophysical data in the archaeological site of Europos (northernGreece). ArchaeologicalProspection, 12, 79–91.

Drahor, M.G., Kurtulmuş, T.Ö., Berge, M.A., Hartmann, M., and Speidel, M.A., (2008).Magnetic imaging and electrical resistivity tomography studies in a Romanmilitary installation found in Satala archaeological site, northeastern Anatolia,Turkey. Journal of Archaeological Science, 35, 259‑271.

Ekinci, Y.L., and Kaya, M.A., (2007). 3D resistivity imaging of buried tombs at the Parionnecropolis (NW Turkey). Journal of Balkan Geophysical Society, 10, 1‑8.

Frish, P., (1983). Die inschri en von Parion, Habelt Edition, Bonn (in German).Hesse, A., Jolivet, A., and Tabbagh, A., (1986). New prospects in shallow depth electrical

surveying for archaeological and pedological applications. Geophysics, 51, 585‑594.

Gaffney, C.F., Gater, J.A., Linford, P., Gaffney, V.L., and White, R., (2000). Large‑scalesystematic fluxgate gradiometry at the Roman city of Wroxeter. ArchaeologicalProspection, 7, 81‑99.

Griffiths, D.H., and Barker, R.D., (1994). Electrical imaging in archaeology. Journal of Archae‑ological Science, 21, 153‑158.

Kasapoğlu, H., (2007). Parion nekropolü mezar tipleri. Doğudan yükselen ışık, arkeolojiyazıları, 481‑521 (in Turkish).

Leaf, W., (1923). Strabo on the Troad. Book XIII, Cap. 1, Cambridge University Press.Leucci, G., Greco, F., De Giorgi, L., and Mauceri, R., (2007). Three‑dimensional image of

seismic refraction tomography and electrical resistivity tomography survey inthe castle of Occhiola (Sicily, Italy). Journal of Archaeological Science, 34, 233‑247.

Loke, M.H., and Barker, R.D., (1996). Rapid least‑squares inversion of apparent resistivitypseudosections using a quasi‑Newton method. Geophysical Prospecting, 44, 131–152.

Loke, M.H., (2001). Tutorial: 2‑D and 3‑D electrical imaging surveys. Penang, Malaysia,Universiti Sains Malaysia, Course notes.

Magie, D., (1950). Roman rule in Asia Minor. Princeton University Press.Mansel, A.M., (1999). Ege ve Yunan Tarihi. Türk Tarih Kurumu Yayınları (in Turkish).Neubauer, W., (1997). A. Eder‑Hinterleitner Resistivity and magnetic of the Roman town

156 YUNUS LEVENT EKINCI et al

Carnuntum, Austria: an example of combined interpretation of prospection data.Archaeological Prospection, 4, 179‑189.

Papadopoulos, N.G., Tsourlos, P., Tsokas, G.N., and Sarris, A., (2006). Two‑dimensionaland Three‑dimensional resistivity imaging in archaeological site investigation.ArchaeologicalProspection, 13, 163–181.

Sasaki, Y., (1992). Resolution of resistivity tomography inferred from numerical simulation.Geophysical Prospecting, 40, 453–464.

Savvaidis, A., Tsokas, G., Liritzis, Y., and Apostolou, M., (1999). The location andmappingof ancient ruins on the castle of Le as (Greece) by resistivity and GPRmethods. Archaeological Prospection, 6, 63–73.

Scollar, I., Weitner, B., and Segeth, K., (1986). Display of archaeological magnetic data.Geophysics, 51, 23‑633.

Tsokas, G.N., Giannopoulos, A., Tsourlos, P., Vargemezis, G., Tealby, J.M., Sarris, A.,Papazachos,C.B., and Savopoulou, T., (1994). A large scale geophysical survey inthe archaeological site of Europos (N. Greece). Journal of Applied Geophysics, 32,85‑98.

Wessel, P., and Smith, W.H.F., (1995). New version of the Generic Mapping Tools released:Eos Trans. AGU. 76, 329.

White, R.M.S., Collins, S., and Loke, M.H., (2003). Resistivity and IP arrays, optimized fordata collection and inversion. Exploration Geophysics, 34, 229‑232.

GEOPHYSICAL IMAGING SURVEY IN THE SOUTH NECROPOLIS AT PARION 157