New Data on the Messapian Necropolis of Monte D’Elia in Alezio (Apulia, Italy) from Topographical and Geophysical Surveys

sensors

Article New Data on the Messapian Necropolis of Monte D’Elia in Alezio (Apulia, Italy) from Topographical and Geophysical Surveys

Giovanni Leucci *, Lara De Giorgi, Immacolata Ditaranto, Francesco Giuri, Ivan Ferrari and Giuseppe Scardozzi

Consiglio Nazionale delle Ricerche—Istituto per i Beni Archeologici e Monumentali, Monteroni road, c/o Campus Universitario, 73100 Lecce, Italy * Correspondence: [email protected]

Received: 27 July 2019; Accepted: 7 August 2019; Published: 9 August 2019

Abstract: The Messapian necropolis of Monte D’Elia is related to one of the most important ancient settlements in the Salento Peninsula (in south Italy). In order to understand the extension and layout of this necropolis in the various periods of its use, a ground-penetrating radar (GPR) prospection was undertaken in some important sample areas by a team of the Institute for Archaeological and Monumental Heritage of the National Research Council of Italy. The analysis of the GPR measurements revealed many anomalies that could be ascribed to archaeological structures (tombs), as well as other anomalies of presumable natural origin or referable to modern features. The data collected were georeferenced in the digital archaeological map of the site and integrated with a virtual reconstruction of the surveyed area.

Keywords: GPR; topographical survey; 3D laser scanning; archaeological map; drone; Messapian necropolis; Alezio

1. Introduction The present study concerns the Messapian necropolis of Monte D’Elia, located about 300 m south of the ancient settlement named Aletium (Aletion). It was inhabited between the late Iron Age and the Roman-Imperial era and lies where modern Alezio stands. The necropolis, used between the 6th and the 2nd century BC, was partially investigated between 1981 and 1985 by archaeological excavations, but the discovered tombs were subsequently buried. Later, a part of the excavated areas was again brought to light and the necropolis became part of an archaeological park. A systematic study of the necropolis does not yet exist: the site, not contain modern overlapping, is very important for the knowledge of Aletium during the Messapian period (in particular between the 6th and the 3rd century BC), when it was one of the main centres of southern Salento. Indeed, the almost complete overlapping of the modern town on the ancient settlement and most of its necropolises greatly limits the reconstruction of the various phases of historical development of Aletion/Aletium, as demonstrated by the recent archaeological map of the site [1] carried out by the Institute for Archaeological and Monumental Heritage of the National Research Council (IBAM-CNR) of Italy. Only part of the grave goods has been studied and neither the extension nor the topographical articulation of the necropolis of Monte D’Elia is known, since archaeological excavations have only partially investigated it. The present research is based on the integration of diﬀerent survey methodologies: (i) the topographical survey of the tombs still visible, also by means of 3D laser scanning techniques; (ii) the production of a large scale archaeological map of the funerary area, starting from ortho-images taken from a drone and thanks to the vectorization and the georeferencing of the old plans drawn in the

Sensors 2019, 19, 3494; doi:10.3390/s19163494 www.mdpi.com/journal/sensors Sensors 2019, 19, 3494 2 of 26 Sensors 2019, 19, x FOR PEER REVIEW 2 of 26

1980s, whichwhich documenteddocumented the the tombs tombs excavated excavated in 1981–1985 in 1981–1985 and nowand buried;now buried; (iii) geophysical (iii) geophysical surveys surveyscarried out carried by means out by of themeans Ground of the Penetrating Ground Pene Radartrating technique Radar both technique on areas both already on aareasﬀected already by the affectedexcavations by the of 1981–1985excavations and of now1981–1985 ﬁlled in,and and now on filled areas in, never and investigated; on areas never (iv) investigated; the geophysical (iv) andthe geophysicalarchaeological and interpretation archaeological of GPR interpretation measurements alsoof GPR thanks measurements to the georeferencing also ofthanks the results to the on georeferencingthe new topographical of the results map. Theon the aim new of this topographi integratedcal study map. wasThe theaim reconstruction of this integrated of the study extension was the of reconstructionthe necropolis andof the its extension topographic of the articulation necropolis during and its the topographic two main phasesarticulation of use, during the Archaic the two period main phasesand the of Late-Classical use, the Archaic/Hellenistic period an age.d the Late-Classical/Hellenistic age.

2. Study Study Area Area The ancient settlement of Aletion/Aletium Aletion/Aletium occupi occupieded the hill today dominated by the medieval Church of S. Maria della Lizza (m 73 a.s.l.) and and the the less less elevated elevated area area located located immediately immediately to to the the east, east, named Raggi (m 65 a.s.l.). The The modern modern Alezio exte extendsnds over over both areas and partly occupies the areas further north. The The site site overlooks overlooks the the fertile fertile coastal coastal plain plain and and is is located located in in an area of adequate height controlling the landing of Gallipoli, just 7 km away (Figure 1 1).). TheThe oldestoldest literaryliterary sourcessources concerningconcerning this ancient settlement date back to the 1st and 2nd centuries AD and documents only its Greek and λ τι ν Latin namesnames AΑλέτιον o /Aletium./Aletium. Therefore, Therefore, the the knowledge knowledge of theof ancientthe ancient phases phases of the of settlement the settlement comes comesexclusively exclusively from archaeological from archaeological data that data offer that a documentation offer a documentation very conditioned very conditioned by the nature by of the naturefindings, of mostlythe findings, random mostly and random not based and on not systematic based on research systematic [2– 6research]. The archaeological [2–6]. The archaeological discoveries, discoveries,carried out between carried theout 19thbetween and 20ththe centuries,19th and 20th and incenturies, particular and from in 1950s,particular were from mostly 1950s, fortuitous were mostlyand took fortuitous place after and the took extension place after of thethe modernextension town of the [1 ].modern Many town of these [1]. findings Many of are these epigraphes findings aredated epigraphes between thedated 6th between and 2nd centuriesthe 6th and BC 2nd and cent Aletionuries/Aletium BC and hasAletion/Aletium returned the highest has returned number the of highestinscriptions number in the of Messapianinscriptions language in the Messapian among the language ancient settlements among the ofancient Salento settlements [7]; most of of them Salento are [7];funerary most of and them many are were funerary found and in themany necropolis were found of Monte in the D’Elia. necropolis of Monte D’Elia.

Figure 1. LocalisationLocalisation of of Alezio Alezio on on DTM DTM of of Salento Salento and al altimetrictimetric model of Alezio Alezio:: A, A, hill hill of of Lizza; Lizza; B, B, Raggi area; C, Monte D’Elia area.

The ancient settlement of Aletion/Aletium Aletion/Aletium was co continuouslyntinuously inhabited from the 7th century BC up to the Middle Ages. It It is is extremely extremely difficult difficult to to read the topography of the inhabited area of the Messapian and Roman times because of the progress progressiveive overlapping of the modern town on the old settlement. The The oldest oldest archaeological archaeological evidence evidence from from the the Alezio Alezio area area dates dates back back to to the the late late Iron Iron Age. Age. It is is likely likely that that at at least least since since the the 7th 7th century century BC BC on on the the hill hill of of Lizza, Lizza, and and probably probably also also in in the the contiguous contiguous Raggi area, there was a village of huts, huts, similar similar to to the the other other settlements settlements that that in in this this period period characterized characterized the Messapia [1]. [1]. Certainly, Certainly, Aletion/Aletium Aletion/Aletium was a se settlementttlement of some importance as early as the 6th century BC, but we do not know if it was aaffectedffected in this period by urban development like other Messapian settlements,settlements, suchsuch asas Cavallino Cavallino and and Oria. Oria. According According to to the the distribution distribution of findingsof findings from from the theArchaic Archaic and Classicaland Classical periods, periods, probably probably the Raggi the areaRaggi was area at least was partlyat least peripheral partly peripheral to the settlement. to the settlement.It perhaps was It perhaps concentrated was concentrated on the hill of on Lizza the and hill could of Lizza occupy and acould surface occupy estimated a surface between estimated 14 and between24 hectares, 14 and but 24 it hectares, cannot be but certainly it cannot defined. be certainly It was defined. characterized It was characterized by nuclei of by houses nuclei with of houses stone with stone foundations alternating with free spaces, partly also used as funerary areas. Several necropolises lay around the inhabited area; the main one of these was located north of the settlement,

Sensors 2019, 19, 3494 3 of 26 foundations alternating with free spaces, partly also used as funerary areas. Several necropolises lay around the inhabited area; the main one of these was located north of the settlement, in the Tafuri area, today completely urbanized. Already at this epoch, the necropolis which extended on the north-western side of Monte D’Elia (65 m a.s.l.), located about 300 m south of the town, was used. The available data for the classical period are very few and insufficient to establish how much the settlement of Aletion/Aletium was affected by the general transformation that during the 5th century BC affected various other Messapian centres, such as Oria and Ugento [5]. However, during the 5th century BC, the necropolis of Monte D’Elia shows a continuity of use. Most of the archaeological data about Aletion/Aletium is dated between the second half of the 4th and the first half of the 3rd century BC [5]. It was hypothesized that in this period the settlement was defended by city walls [4,8], of which no certain remains are preserved. On the basis of traces visible in the aerial photographs of the 1940s, a perimeter of 3.400 m was hypothetically reconstructed for these walls, enclosing a surface of about 67 hectares and including both the hill of Lizza and the Raggi area [1]. In the lack of data on housing, the indirect evidence to the flowering of the settlement in this phase comes almost exclusively from the necropolises that surrounded it and among which the necropolis of Monte D’Elia stands out, as in previous phases. The ancient settlement of Alezio seems to have maintained a certain importance even after the Roman conquest of Salento, marked by the triumph over the Calabri and Sallentini in 266 BC. According to the available archaeological evidence, in this phase the settlement seems to be decreasing, concentrating on the hill of Lizza and in the central-western sector of the Raggi area. Between the late Republican age and the Imperial period, the settlement seems to lose a large part of its urban character to undergo a strong process of ruralization, marked by the presence of facilities for the transformation of agricultural products, such as olive oil and wine. The settlement survived until the Middle Ages with a progressive abandonment of housing structures and depopulation.

3. Topographical and Archaeological Survey

3.1. Topographical and 3D Survey The topographical survey and the graphical representation of the Messapian necropolis of Monte d’Elia contributed to verify and to correct the survey made during the archaeological excavations in the 1980s. It has allowed us to update the cartography for a critical reading of the site: a useful tool for the investigation of the subsoil introduced in this paper. Different indirect survey techniques have been used since the use of a single type of measurement does not give good results in terms of geometric accuracy, portability, automatism and photorealism. The sensors used for the survey use light radiation; a further distinction can be made depending on the nature of the light used to make the measurement. If natural light is used, measuring instruments are defined as “passive” (photogrammetric technique, Image-Based, etc.); if the light is generated in the measurement process, we are talking about “active sensors” (laser scanners, radars, total stations, etc.). The integration between different systems of three-dimensional relief represents a central topic of research in different disciplinary fields, within which new methods are being studied to integrate various techniques in an automatic or semi-automatic manner, enhancing their potential [9]. Both active and passive sensors show a high level of complementarity. Digital photogrammetry is very useful when there is a limited permanence in the place of the surveyors or if the artefact can be described exhaustively by control points and lines. On the contrary, laser scanner requires more time in data collecting and in the post-processing phase from merging of the cloud points up to the processing of the polygonal mesh.This technique clearly offers very precise data not achievable with traditional photogrammetry. The use of different surveying instruments offers metric data characterized by multiresolution and allows to achieve an exhaustive result with the different representation of all geometries [10]. Sensors 2019, 19, 3494 4 of 26

The archaeological park of Monte D’Elia has a total area of about 11,000 square meters. Inside it, only a part of the necropolis (corresponding to a surface of about 636 square meters) is currently visible. Sensors 2019, 19, x FOR PEER REVIEW 4 of 26 For a fast survey of the entire park, aerial digital photogrammetry was performed using an Unmanned Aerial Vehicle (UAV), while for the necropolis area, unfortunately covered by a metal roof, the survey roof, the survey using close range digital photogrammetry has been implemented with e laser scanner using close range digital photogrammetry has been implemented with e laser scanner data (Figure2). data (Figure 2).

FigureFigure 2. 2. TheThe 3D 3Dsurvey survey of the of archaeological the archaeological site of Monte site of d’Elia Monte created d’Elia with created the Agisoft with the Photoscan Agisoft software.Photoscan software.

TheThe UAVUAV airframeairframe used used was was a Scrabble a Scrabble 4HSE 4HSE quadcopter quadcopter designed designed and manufactured and manufactured by Italdron by Italdron(Ravenna, (Ravenna, Italy) and Italy) equipped and equipped with a high with resolution a high resolution digital camera digital (DMC-GH4, camera (DMC-GH4, 16 Mpx, focal 16 length: Mpx, 28.0 mm; sensor resolution. 4608 2592 pixels, Panasonic, (Milano, Italy), fixed on a 3-axis gimbal. focal length: 28.0 mm; sensor resolution.× 4608 × 2592 pixels, Panasonic, (Milano, Italy), fixed on a 3- axisThe gimbal. maximum The weight maximum was 12weight kg, allowing was 12 kg, a maximum allowing a flight maximum endurance flight limit endurance of 10 min. limit To of cover 10 min. the Towhole cover area, the flights whole were area, performed flights were with aperformed nadir orientation with a of nadir the camera. orientation Three of additional the camera. flights Three with additionalmanual piloting flights with with a 45manual◦ camera piloting were carriedwith a 45° out camera to collect were oblique carried aerial out images. to collect The oblique redundant aerial set images.of images The acquired redundant at the set selected of images timing acquired facilitates at the the selected SfM approach. timing facilitates In total, 706the aerialSfM approach. photos have In total,been 706 taken aerial using photos a constant have been shooting taken distance using a of constant about 5 shooting m, at a midrange distance distanceof about from5 m, at the a midrange ground of distance30 m. from the ground of 30 m. TheThe entire entire archaeological archaeological park park was was covered covered with with a a sufficient sufficient image image overlap overlap (about (about 70–80%), 70–80%), essentialessential in order to getget thethe trackingtracking pointspoints in in space space and and their their resulting resulting 3D 3D position. position. The The processing processing of ofthe the images images by by Agisoft Agisoft Photoscan Photoscan software software (v. (v. 1.4.0.5076, 1.4.0.5076, Agisoft, Agisoft, St.St. Petersburg,Petersburg, Russia) led to the the alignmentalignment of of the the photos photos with with a a mini minimummum margin margin of of error error (0.6/2.0 (0.6/2.0 pixel) pixel) and and to to the the creation creation of of a a dense dense cloudcloud of of about about 19 19 million million points. points. The The model model obtained obtained was was georeferenced georeferenced through through four four ground ground control control pointspoints (GCPs) (GCPs) located located on on the the corners corners of of the the park park and and collected collected with with a a total station, to to guarantee its its correctcorrect scaling scaling and and orientation orientation [11–14]. [11–14]. InIn the the same same Photoscan Photoscan project project a a second second “chunk” “chunk” wa wass detected detected in in which which the the images images of of the the second second photographicphotographic surveysurvey ofof the the area area not not covered covered from from the the UAV UAV (due (due to the to metal the metal roof protectingroof protecting the visible the visibletombs) tombs) were aligned. were aligned. Still with Still the with same the camerasame camera and by and means by means of a telescopic of a telescopic pole, pole, 627 photos 627 photos were weretaken. taken. The The alignment alignment of all of photosall photos has has a margin a margin of of error error 0.3 0.3/1.4/1.4 pixel pixel and and producedproduced aa finalfinal digital model of 8 million polygons with a texture of 15,000 15,000 px resolution (Figure3). Both “chunks” model of 8 million polygons with a texture of 15,000 ×× 15,000 px resolution (Figure 3). Both “chunks” werewere merged by placing markers in in the the common common poin points,ts, thus thus obtaining obtaining a a single single 3D 3D model model used used as as a a basisbasis for for placing placing geophysical geophysical pr prospectingospecting results (Figure 4 4).).

Sensors 2019, 19, 3494 5 of 26 Sensors 2019, 19, x FOR PEER REVIEW 5 of 26 Sensors 2019, 19, x FOR PEER REVIEW 5 of 26

FigureFigure 3. The3. The 3D 3D survey survey of of the the tombs tombs currentlycurrently visible at at necropolis necropolis of of Monte Monte D’Elia D’Elia created created with with the the AgisoftAgisoft Photoscan Photoscan software. software.

Figure 4. The 3D survey: (A), the digital aerial photogrammetry of the archaeological site; (B), the FigureFigure 4. The 4. The 3D 3D survey: survey: (A ),(A the), the digital digital aerial aerial photogrammetry photogrammetr ofy theof the archaeological archaeological site; site; (B ),(B the), the aerial aerial survey merged with the digital close range photogrammetry of the tombs current visible under surveyaerial merged survey withmerged the with digital the di closegital rangeclose range photogrammetry photogrammetry of the of the tombs tombs current current visible visible under under the the metal cover; (C), CAD plan. metal cover; (C), CAD plan.

Sensors 2019, 19, 3494 6 of 26

After analysing the morphological characteristics of the site and the presence of several sarcophagi and chests tombs, we have set up scanning stations at an average distance of about 5 to 7 m, between each other, in order to have a good coverage of all surfaces. The survey was performed with a P20 laser scanner (Leica, Milano, Italy) using an accuracy of 6 mm on a dome of 10 m radius [15–18]. The matching of point clouds was done manually. From them, a three-dimensional model was obtained with a mesh of about 7 million polygons with a resolution of 10 mm. Plan and sections of the area have been processed (Figure5) starting from the scaled model. Sensors 2019, 19, x FOR PEER REVIEW 6 of 26

FigureFigure 5. Laser 5. Laser scanner scanner survey survey of of the the archaeological archaeological site site of ofMonte Monte D’Elia: D’Elia: plan and plan sections. and sections.

Sensors 2019, 19, x FOR PEER REVIEW 7 of 26 with a P20 laser scanner (Leica, Milano, Italy) using an accuracy of 6 mm on a dome of 10 m radius [15–18].Sensors 2019 The, 19 ,matching 3494 of point clouds was done manually. From them, a three-dimensional model7 of 26 was obtained with a mesh of about 7 million polygons with a resolution of 10 mm. Plan and sections of the area have been processed (Figure 5) starting from the scaled model. Finally, the work flow planned and the data obtained allowed us to manage all information within Finally, the work flow planned and the data obtained allowed us to manage all information a single 3D work space, and to produce two digital models aimed at a quickly providing all data about within a single 3D work space, and to produce two digital models aimed at a quickly providing all the geometry and the colour of the analysed scenario. data about the geometry and the colour of the analysed scenario. The integration between these techniques demonstrates how single instruments are characterized The integration between these techniques demonstrates how single instruments are by a level of complementarity that makes an integrated system more powerful and flexible, able to characterized by a level of complementarity that makes an integrated system more powerful and provide a much better result in absolute terms and able itself to adapt to the single morphological flexible, able to provide a much better result in absolute terms and able itself to adapt to the single features of the different objects contained in the detected scene. The complementarity of the techniques morphological features of the different objects contained in the detected scene. The complementarity optimize the acquisition and modelling process, exploiting the maximum potential of the single of the techniques optimize the acquisition and modelling process, exploiting the maximum potential instrument. The result achieved is a digital cognitive model, which is a repository of information; of the single instrument. The result achieved is a digital cognitive model, which is a repository of therefore, search queries can be carried out at various levels (Figure4B). information; therefore, search queries can be carried out at various levels (Figure 4B). 3.2. Archaeological Map 3.2. Archaeological Map Between 1981 and 1985, archaeological excavations were carried out in the Messapian necropolis of MonteBetween D’Elia. 1981 The and investigations 1985, archaeological involved excavations a large sector were of thecarried necropolis out in the and Messapian they brought necropolis to light ofabout Monte 60 tombs,D’Elia. whoseThe investigations grave goods involved attest that a thelarge use sector of the of funerary the necropolis area, related and they to thebrought settlement to light of Aletionabout 60/Aletium, tombs, whose was between grave goods 6th and attest 2nd that century the use BC of (Figure the funerary6)[ 1,3 –area,5,19 –related26]. The to funerarythe settlement types ofvary Aletion/Aletium, from the pit graves was withbetween cover 6t consistingh and 2nd ofcentury a slab ofBC local (Figure calcarenite, 6) [1,3–5,19–26]. dating in The the Archaicfunerary period, types varyto the from monolithic the pit sarcophagigraves with and cover to the consisting large chests of a madeslab of of local more calcarenite, calcarenite dating slabs, coveredin the Archaic by one period,or more to slabs, the monolithic and in use sarcophagi between 5th and and to 3rd the century large chests BC. Various made of not more inscribed calcarenite cippi slabs, in calcarenite covered byhave one also or beenmore found,slabs, and associated in use between with both 5th Archaic and 3rd and century Hellenistic BC. Various tombs [24not,27 inscribed,28]. They cippi are in of calcareniteparallelepiped have shape also been (exemplifying found, associated measures: with 34 both22 Archaic cm, height and 65 Hellenistic cm). tombs [24,27,28]. They are of parallelepiped shape (exemplifying measures:× 34 × 22 cm, height 65 cm).

Figure 6. PlanPlan of of the the excavated excavated area of Monte D’Elia necropolis.

The discovered discovered tombs tombs were were buried buried in in1989 1989 for forconservation conservation reasons reasons and andlargely largely brought brought to light to againlight againin 2004, in 2004,when when the area the area became became an anarchaeological archaeological park. park. The The available available documentation documentation is incomplete,is incomplete, the the published published works works are arefew fewand andthe plans the plans of the of excavated the excavated areas areas are inconsistent are inconsistent with witheach other each otherand without and without georeferencing. georeferencing. In particular In particular,, only two only general two plans general are plans available: are available: the first, drawnthe first, by drawn Duma by and Duma Zingariello, and Zingariello, and preserved and preserved in the inarchive the archive of the of Superintendent the Superintendent for the for Archaeologicalthe Archaeological Heritage Heritage of Puglia of Puglia Region, Region, docume documentsnts the excavations the excavations of 1981–1982; of 1981–1982; the second the second one, drawnone, drawn by Danese by Danese in 1987, in 1987, shows shows both both the the excavations excavations of of1981–1982 1981–1982 (but (but with with some some differences differences compared to to the the Duma-Zingariello Duma-Zingariello map) map) and and those those of of 1985 1985 (this (this last last survey survey is iskept kept in inthe the archive archive of theof the Municipality Municipality of Alezio of Alezio and and was was published published in 1981 in1981 [3] (Figure [3] (Figure 2) and2) [5] and (Figure [ 5] (Figure 5) and,5) then, and, then,in [1] in [1] (Figure4, with changes). The currently visible area of the necropolis includes most of the tombs brought to light in 1981 and three of those excavated in 1982. The execution of a new plan of the visible Sensors 2019, 19, 3494 8 of 26 tombs, georeferenced in the new topographical map of the site, allowed evaluation of the accuracy of the reliefs of Duma-Zingariello and Danese (which in previous publications was always oriented to the south). In particular, the Duma-Zingariello drawing (which also shows the official denomination and numbering of the graves unearthed in 1981 and part of those discovered in 1982) was very faithful to the real situation and therefore considered reliable even for the tombs no longer in sight. The plan of Danese, on the other hand, was less precise, but it is the only documentation available for the tombs unearthed in 1985. Furthermore, the comparison of the plan of Duma-Zingariello with the current situation showed that at the time of the reopening of the excavations, in 2004, some of the Archaic tombs brought to light in 1981, which had more fragile structures, were destroyed, while others were slightly moved from their original position. Moreover, none of the plans show the heights above ground level or sea level are never shown. The production of a new archaeological map of the necropolis, including all the discovered tombs, and the contextualization of data from grave goods and epigraphic documentation allowed some considerations and observations about the funerary area of Monte D’Elia and its organization. The excavation of 1981 is undoubtedly the best documented, thanks also to a preliminary study of the grave goods that allowed establishing the general chronology of the necropolis. This excavation involved an area extending 35 m in the east-west direction and 17 m in the north-south direction and led to the discovery of 37 tombs. The tombs referring to the Archaic period according to typology and grave goods, and datable between the 6th and the beginning of the 5th century BC, are at least ten (nos. 8, 10, 11, 12, 18, 19, 23, 24, 26, 29); we can also probably add the tombs nos. 28, 32, 34, characterized by the same typology. With the exception of the tomb no. 19, which is a sarcophagus (1 1.65 m) and is × dated between the end of the 6th and the beginning of the 5th century BC, the others tombs are all pit graves or excavated in the calcareous bench (exemplifying measures: 60 80 cm) and covered by a × calcarenite slab (max size 0.85 1.25 m). The deceased appear in a crouched position and the grave × goods include a small number of objects [1,3,19,24,25,29]. They generally comprising a small olla and a one handle jug, sometimes accompanied by a one handle uncoloured cup, all of local production, for female burials (nos. 12, 18, 19), and a krater associated with an imported drinking cup (Ionic cups type B2 or imitations of Attic cups Bloesch C) and other vessels of local production (such as jugs or small jugs ladle, lekanai, and stamnoi), for the male tombs (nos. 8, 10, 11, 23, 24, 26, 29). In particular, the more ancient male burials (nos. 10, 11, 24, 29), datable to the second half of the 6th century BC, are characterized by the presence of globular Messapian kraters with mushroom-shaped handles accompanied by Ionic cups of type B2, likely produced at Metaponto. The more recent tombs (nos. 23 and 26), dating between the end of the 6th and the beginning of the 5th century BC, are characterized instead by the presence of a column-krater that constitutes a local imitation of Greek models and is always associated with imitations of the Attic cups like Bloesch C, referring both to the production of Taranto that to the Corinthian colonies of the Adriatic Sea [29]. A few metallic materials were found, mostly related to fibulae. The new relief of the still visible tombs has highlighted that the nos. 10, 26, 28, 29, 30 and 34, documented in the Duma-Zingariello plan and in some photos collected in 1981 [3] (Figure3) and [ 5] (Figure6), were unfortunately destroyed during the excavation of 2004, while the remains of the nos. 12 and 24 have been slightly displaced with respect to the original position. The location of tomb no. 8, today not preserved and absent in the Duma-Zingariello plan, is uncertain. The most consistent group of tombs dated between the 5th and 3rd century BC and it consists of sarcophagi (nos. 5, 9, 13, 15, 16, 17, 21, 25, 33) and chests made of slabs (nos. 1, 2, 3, 4, 6, 7, 14, 20, 22, 31, 35, 36, 37). Among the first group, sarcophagus no. 5 (m 1.65 0.65, height m 0.45) was × found to the east of the investigated area, close to a cippus, and was transferred in the garden of the Museo Civico Messapico of Alezio. Other three tombs (nos. 16, 21, 33) are very small (for example 50 80 cm). Among chests of slabs (exemplifying measures: lengths 1.74–2.97 m, widths 0.86–1.33 m, × heights 0.70–1.20 m), often characterized by cover slabs provided with recesses for the handle, eight (nos. 1, 2, 3, 4, 6, 20, 22, 31) have Messapian inscriptions engraved on the inner faces of one side slab Sensors 2019, 19, 3494 9 of 26 and bearing the names of the owners of the tombs themselves. They are datable between the second half of the 5th and the end of the 3rd-beginnings of the 2nd century BC [1,7,30–33]. About this group of tombs, four (nos. 1, 2, 3, 6) were transferred in the garden of the Museo Civico Messapico of Alezio and belonged to a nucleus located in the north-eastern sector of the investigated area, including also tomb no. 4, today buried. The grave goods of these tombs have not yet been carefully studied (see notes in [1,4,5,25]). In some cases, they go back to the 2nd century BC (for example the tombs nos. 14 and 31) suggesting their reuse or use over several generations. The grave goods generally include local trozzelle and lekanai of local production, black-painted pottery (such as lekythoi, patere, skyphoi, oil lamps), various types of uncoloured vases, balsamari, as well as small female statues sitting on a throne or standing (as in the tombs nos. 3, 6, 22), egg shells (as in the tombs nos. 3, 6, 15) and few metallic objects (such as an iron brooch and a bronze earring from tomb no. 3). The excavation carried out in 1982 involved an area of almost 50 m wide in the east-west direction, located just north of that investigated in 1981. Regular squares of about 4 4 m were excavated, and in × some cases were further enlarged. At least 11 tombs (sarcophagi and chests made of slabs) were found (nos. 1/82–11/82); three others tombs may be added (nos. 12/82–14/82), which are drawn in the Danese plan but not in Duma-Zingariello plan. Only sarcophagi nos. 6/82–10/82 are currently visible and they are located at the north-western end of the tombs brought to light in 1981. The rest is buried, as well as all the tombs (at least nine), always sarcophagi and chest of slabs, brought to light in 1985 to the east of the area investigated in 1982, where an excavation of about 11 8 m was carried out. Some slabs of the × north-eastern excavated sector (tombs nos. 1/85–9/85) were removed (some of these slabs are piled in the vicinity), but we don’t have further data about the tombs that are still in situ and buried. It is possible that also in the northern excavated sector (nos. 1/82–12/82) some tombs were removed. Unlike the excavation of 1982 which did not allow the recovery of new inscriptions, three chest tombs discovered in 1985 (example measurements: lengths 1.25–2.23 m, widths 0.80–0.88, heights 0.60) returned three Messapian epigraphs of the 4th–3rd century BC [1,7,34]. The inscriptions, while awaiting the study of the grave goods, allow us to date at least part of these tombs to the Late-Classical and Hellenistic age. It should be emphasized that there are no data about the presence of Archaic tombs, even in the areas investigated in 1982 and 1985. Overall, the graves unearthed between 1981 and 1985 occupy an area extending 25 m in north-south direction and about 75 m in east-west direction (Figure7). An isolated tomb was also brought to light approximately 25 m south-west of the area excavated in 1981. The necropolis was organised by terracing a natural slope slightly downhill towards the west, with the graves were positioned on different levels. In the westernmost sector, characterized by a very thick layer of earth above the rocky bank, thanks to the still visible graves it is possible to see that the tombs were found starting from a depth of 70–80 cm and that the deepest ones lie at 1.5/ 2 m compared to the ground level. In the western sector − − the tombs are all oriented in a north-west/south-east direction. It is possible that this orientation was influenced by an ancient road paved with compacted small stones, identified, only in a short section, in 2004 about 10 m north-west of the funerary area. The road, which perhaps ran between the main nucleus of the tombs brought to light in 1981 and the group consisting of tombs nos. 1–4 and 6, could be related to the north-west/south-east oriented road documented in 1999 by geophysical surveys, still unpublished, carried out by the University of Sydney [6,25]. Lastly, as we proceed towards the east, the tombs take on a marked north-south or east-west orientation, as is evident especially in the eastern sector excavated in 1985. The orientation of this area could be related to a different road network: indeed, it may be interesting to remember that the modern road bounding the eastern side of the archaeological park is identified with the survival of the so-called Via Sallentina [35], the main road that connected the ancient settlements of Aletium and Uxentum (today Ugento). Sensors 2019, 19, x FOR PEER REVIEW 9 of 26 bearing the names of the owners of the tombs themselves. They are datable between the second half of the 5th and the end of the 3rd-beginnings of the 2nd century BC [1,7,30–33]. About this group of tombs, four (nos. 1, 2, 3, 6) were transferred in the garden of the Museo Civico Messapico of Alezio and belonged to a nucleus located in the north-eastern sector of the investigated area, including also tomb no. 4, today buried. The grave goods of these tombs have not yet been carefully studied (see notes in [1,4,5,25]). In some cases, they go back to the 2nd century BC (for example the tombs nos. 14 and 31) suggesting their reuse or use over several generations. The grave goods generally include local trozzelle and lekanai of local production, black-painted pottery (such as lekythoi, patere, skyphoi, oil lamps), various types of uncoloured vases, balsamari, as well as small female statues sitting on a throne or standing (as in the tombs nos. 3, 6, 22), egg shells (as in the tombs nos. 3, 6, 15) and few metallic objects (such as an iron brooch and a bronze earring from tomb no. 3). The excavation carried out in 1982 involved an area of almost 50 m wide in the east-west direction, located just north of that investigated in 1981. Regular squares of about 4 × 4 m were excavated, and in some cases were further enlarged. At least 11 tombs (sarcophagi and chests made of slabs) were found (nos. 1/82–11/82); three others tombs may be added (nos. 12/82–14/82), which are drawn in the Danese plan but not in Duma-Zingariello plan. Only sarcophagi nos. 6/82–10/82 are currently visible and they are located at the north-western end of the tombs brought to light in 1981. The rest is buried, as well as all the tombs (at least nine), always sarcophagi and chest of slabs, brought to light in 1985 to the east of the area investigated in 1982, where an excavation of about 11 × 8 m was carried out. Some slabs of the north-eastern excavated sector (tombs nos. 1/85–9/85) were removed (some of these slabs are piled in the vicinity), but we don’t have further data about the tombs that are still in situ and buried. It is possible that also in the northern excavated sector (nos. 1/82–12/82) some tombs were removed. Unlike the excavation of 1982 which did not allow the recovery of new inscriptions, three chest tombs discovered in 1985 (example measurements: lengths 1.25–2.23 m, widths 0.80–0.88, heights 0.60) returned three Messapian epigraphs of the 4th–3rd century BC [1,7,34]. The inscriptions, while awaiting the study of the grave goods, allow us to date at least part of these tombs to the Late- ClassicalSensors 2019 and, 19, 3494Hellenistic age. It should be emphasized that there are no data about the presence10 of of 26 Archaic tombs, even in the areas investigated in 1982 and 1985.

Figure 7. The new map of the funerary area of Monte D’Elia on which the seven investigated areas are highlighted.

4. GeophysicalOverall, the Survey graves unearthed between 1981 and 1985 occupy an area extending 25 m in north- southA direction wide range and of about geophysical 75 m in methodseast-west are direction applied (Figure in archaeology 7). An isolated for obtaining tomb was high-resolution also brought images of the subsurface. The geophysical method used in this study is based on the detection of variations in the electromagnetic properties of the subsoil and the use of these data to identify artefacts and distinguish between these and natural variations in the soil. Seven areas, labelled respectively A, B, C, D, E, F, and G were considered for GPR survey (Figure7). It was necessary to choose seven di ﬀerent areas due to the current conditions of the archaeological park: in fact, the presence of numerous obstacles, such as walls, roads, trees, and light poles, did not allow us to cover the overall surface of the site.

4.1. GPR Data Acquisition and Analysis The GPR survey was carried out with an IDS Ris Hi Mod system (IDS, Pisa, Italy) using a 200–600 MHz dual band antenna. Data were acquired in continuous mode on the overall surface of the seven investigated areas, along 0.5 m-spaced survey lines, using 512 samples per trace, 80 ns time window for 600 MHz antenna and 160 ns time window for 200 MHz antenna, and a manual time-varying gain function. Transect spacing should be less than one half the wavelength of possible reflections returned from the smallest target to be mapped [36–38]. The data were subsequently processed using standard two-dimensional processing techniques by means of the GPR-Slice Version 7.0 software [39]. The processing flow-chart consists of the following steps: (i) header editing for inserting the geometrical information; (ii) frequency filtering; (iii) manual gain, to adjust the acquisition gain function and enhance the visibility of deeper anomalies; (iv) customized background removal to attenuate the horizontal banding in the deeper part of the sections (ringing), performed by subtracting in different time ranges a ‘local’ average noise trace estimated from suitably selected time-distance windows with low signal content (this local subtraction procedure was necessary to avoid artefacts created by the classic subtraction of a ‘global’ average trace estimated from the entire section, due to the presence of zones with a very strong signal); (v) estimation of the average electromagnetic wave velocity by hyperbola fitting; (vi) Kirchhoff migration, using a constant average velocity value of 0.098 m/ns. The migrated data were subsequently merged together into Sensors 2019, 19, 3494 11 of 26 three-dimensional volumes and visualized in various ways in order to enhance the spatial correlations of anomalies of interest. A way of obtaining visually useful maps for understanding the plan distribution of reflection amplitudes within specific time intervals is the creation of horizontal time slices. These are maps on which the reflection amplitudes have been projected at a specified time (or depth), with a selected Sensors 2019, 19, x FOR PEER REVIEW 11 of 26 time interval [40]. In a graphic method developed by Goodman et al. [41], termed ‘overlay analysis’, thethe strongeststrongest andand weakestweakest reflectorsreflectors atat thethe depthdepth ofof eacheach sliceslice areare assignedassigned specificspecificcolours. colours. ThisThis techniquetechnique allowsallows thethe linkagelinkage ofof structuresstructures buriedburied atat didifferentfferent depths.depths. This represents an improvement inin imagingimaging because because subtle subtle features features that arethat indistinguishable are indistinguishable on radargrams on radargrams can be seen can and be interpreted seen and ininterpreted a more easily. in a Inmore the easily. present In work the present the time-slice work the technique time-slice has technique been used has to display been used the amplitudeto display variationsthe amplitude within variations consecutive within time consecutive windows time of width windows∆t = of5 ns.width Moreover, Δt = 5 ns. the Moreover, highest amplitudes the highest wereamplitudes rendered were into rendered an iso-surface into an [iso-surface36,42–44]. Three-dimensional[36,42–44]. Three-dimensional amplitude iso-surfaceamplitude iso-surface rendering displaysrendering amplitudes displays amplitudes of equal value of equal in the value GPR in study the GPR volume. study Shading volume. is usuallyShading used is usually to illuminate used to theseilluminate surfaces, these giving surfaces, the appearancegiving the appearance of real archaeological of real archaeological structures. structures. In this case In the this threshold case the calibrationthreshold calibration is a very delicate is a very task delicate in order task to in obtain order usefulto obtain results. useful In results. order toIn defineorder theto define depth the of archaeologicaldepth of archaeological remains the remains electromagnetic the electromagneti (EM) wavec (EM) velocity, wave using velocity, the characteristic using the characteristic hyperbolic shapehyperbolic of a reflection shape of froma reflection a point from source a (dipointffraction source hyperbola), (diffraction was hyperbola), used. Only was the used. 600 MHz Only antenna the 600 resultsMHz antenna were considered results were due considered to the fact due that to no the additional fact that informationno additional were information visible on were the visible 200 MHz on antennathe 200 MHz results antenna (see Figure results8b). (see Figure 8b).

FigureFigure 8.8. Area A: (a) 600 MHz antenna: reflection reﬂection events related probably toto a modern pipepipe (4),(4), tombs (1,(1, 2)2) andand toto thethe bedrock,bedrock, atat aa depthdepth betweenbetween 0.750.75 toto 1.351.35 m;m; ((bb)) 200200 MHzMHz antenna:antenna: nono moremore evidencedevidenced werewere foundfound belowbelow 130130 ns.ns.

4.2.4.2. Geophysical Anomalies

4.2.1. Area A 4.2.1. Area A Area A was about 26 19 m. In this area an electromagnetic energy penetration depth of 40–50 ns Area A was about 26× × 19 m. In this area an electromagnetic energy penetration depth of 40–50 was found. Figure8 shows the processed radargram related to the 14th profile and to the 600 MHz ns was found. Figure 8 shows the processed radargram related to the 14th profile and to the 600 MHz antenna. It shows several hyperbolic shaped reflection events. The first reflection event, labelled 4, antenna. It shows several hyperbolic shaped reflection events. The first reflection event, labelled 4, at at two-way travel time window between 15 and 20 ns is clearly visible. Its size is about 0.2 m and the two-way travel time window between 15 and 20 ns is clearly visible. Its size is about 0.2 m and the depth is between 0.68 and 0.90 m (with an average electromagnetic wave velocity of 0.09 m/ns). This depth is between 0.68 and 0.90 m (with an average electromagnetic wave velocity of 0.09 m/ns). This could be related to a modern pipe. On each of the GPR records the lowest (dashed yellow) continuous could be related to a modern pipe. On each of the GPR records the lowest (dashed yellow) continuous and slightly undulating reflector appears strong and irregular and reaches a maximum depth below and slightly undulating reflector appears strong and irregular and reaches a maximum depth below the ground surface ranging from 0.75 to 1.35 m. This event could be related to the shallow bedrock visible also in the old excavations (Figure 6). The reflection events labelled 1 and 2 are visible at depth from about 0.45 m to 0.60 m. The size (about 0.80 m) and shape suggest that they are probably related to the presence of filled tombs (Figure 8). In order to identify the depth evolution of buried structures, including their size, shape and location, time slices using the overlay analysis [41,45] were built (Figure 9). The time slices show the normalized amplitude using a range defined by blue as zero and red as 1. In the slices ranging from 0.34 m to 0.58 m depth, relatively high-amplitude alignments (labelled 1, 2, 5 and 6 respectively) are

Sensors 2019, 19, 3494 12 of 26 the ground surface ranging from 0.75 to 1.35 m. This event could be related to the shallow bedrock visible also in the old excavations (Figure6). The reflection events labelled 1 and 2 are visible at depth from about 0.45 m to 0.60 m. The size (about 0.80 m) and shape suggest that they are probably related to the presence of filled tombs (Figure8). In order to identify the depth evolution of buried structures, including their size, shape and location, time slices using the overlay analysis [41,45] were built (Figure9). The time slices show the normalized amplitude using a range defined by blue as zero and red as 1. In the slices ranging Sensors 2019, 19, x FOR PEER REVIEW 12 of 26 from 0.34 m to 0.58 m depth, relatively high-amplitude alignments (labelled 1, 2, 5 and 6 respectively) clearlyare clearly visible. visible. These These correspond correspond to the to anomalies the anomalies labelled labelled 1 and 2 1 in and the 2 radargram in the radargram (Figure (Figure 8). In the8). timeIn the slices time (Figure slices (Figure 9) ranging9) ranging from 0.70 from m 0.70 to 0.90 m to m 0.90 depth m depththe dashed the dashed dark line dark highlights line highlights a high- a amplitudehigh-amplitude anomaly anomaly (labelled (labelled 4); in 4); the in same the same area, area, the thelatest latest two two time time slices slices ranging ranging from from 1.39m 1.39m to to 1.8m1.8m depth depth show show other other high amplitude events (labelled 3 and 7).

FigureFigure 9. 9. TimeTime slices slices of of Area Area A: A: the the most most significant signiﬁcant anomalies, anomalies, probably probably related related to to tombs, tombs, are are highlightedhighlighted at at depth depth of of 34–58 34–58 cm cm (1, (1, 2, 2, 5, 5, 6) 6) and and 139–180 139–180 cm cm (3, (3, 7). A A modern modern pipe pipe is is at at 70–93 70–93 cm cm (4). (4).

4.2.24.2.2. Area Area B B Area B was about 6 30 m in size. The time slice analysis (Figure 10) show several high Area B was about 6 × 30 m in size. The time slice analysis (Figure 10) show several high electromagneticelectromagnetic energy amplitude events. In In the the slices slices ranging ranging from from 0.33 0.33 m m to to 0.56 0.56 m m depth, depth, relatively relatively high-amplitudehigh-amplitude alignments alignments (labelle (labelledd 12, 12, and and 11, 11, respectively) respectively) are clea clearlyrly visible. These These by by their their shape shape (rectangular) and dimensions (about 0.8 m 1.8 m) could be related to the presence of tombs. In the (rectangular) and dimensions (about 0.8 m × 1.8 m) could be related to the presence of tombs. In the timetime slices slices ranging ranging from from 0.50 0.50 m m to to 0.70 0.70 m m depth depth the the dashed dashed dark dark line line highlights highlights a a high-amplitude high-amplitude anomalyanomaly (labelled (labelled 13 13 and and 14) 14) that that seems seems to to indicate indicate the the presence presence of of archaeological archaeological features features (walls); (walls); in in thethe same same slice slice the the anomaly anomaly labelled labelled 10 10 is is also also visible. visible. In In the the same same area, area, deeper deeper time time slice slice ranging ranging from from 1.011.01 m m to to 1.24 m depth shows other high amplitude events (labelled 8 8 and 9). Also Also in in this this case case the the alignmentalignment with with the the other other excavated excavated tombs tombs allows allows one to interpret these as probable tombs.

4.2.3. Area C Area C was about 15 × 13 m. The time slice analysis (Figure 11) show several high electromagnetic energy amplitude events labelled 15, 16, 17 and 18, respectively. The shape, the dimensions and the alignment with the other excavated tombs allow one to interpret these as probable tombs.

Sensors 2019, 19, 3494 13 of 26 Sensors 2019, 19, x FOR PEER REVIEW 13 of 26

Sensors 2019, 19, x FOR PEER REVIEW 13 of 26

Figure 10. TimeTime slices slices of of Area Area B: B: the most significan signiﬁcantt anomalies, probably related to tombs, are highlighted at depth of 33–73 cm (10, 11, 12) and 101–124 cm (8, 9). A A modern modern pipe pipe is is at at 70–93 70–93 cm cm (4). (4). Linear anomalies (13, 14) are visiblevisible at 50–73 cm (square structure?).

4.2.3. Area C

Area C was about 15 13 m. The time slice analysis (Figure 11) show several high electromagnetic Figure 10. Time slices× of Area B: the most significant anomalies, probably related to tombs, are energy amplitude events labelled 15, 16, 17 and 18, respectively. The shape, the dimensions and the highlighted at depth of 33–73 cm (10, 11, 12) and 101–124 cm (8, 9). A modern pipe is at 70–93 cm (4). alignmentLinear with anomalies the other (13, 14) excavated are visible tombs at 50–73 allow cm one(square to interpret structure?). these as probable tombs.

Figure 11. Time slices of Area C: the most significant anomalies, probably related to tombs, are highlighted at depth of 28–60 cm (17, 18) and 69–130 cm (15, 16).

4.2.4. Area D Area D was about 13 × 20 m. The time slice analysis (Figure 12) show several high electromagneticFigureFigure 11.11. Time Timeenergy slicesslices amplitude ofof AreaArea C:C: events thethe mostmost labelled signiﬁcantsignificant 20, 21, anomalies,anomalies, 22 and probablyprobably23, respectively. relatedrelated toto The tombs,tombs, shape, areare the dimensionshighlightedhighlighted and at atthe depthdepth alignment of of 28–60 28–60 cm cmwith (17, (17, 18)the 18) and andother 69–130 69–130 excavated cm cm (15, (15, 16). 16).tombs allow one to interpret these as probable tombs. 4.2.4.4.2.4. AreaArea DD 4.2.5. Area E AreaArea D D was was about about 13 2013 m.× The20 m. time The slice time analysis slic (Figuree analysis 12) show(Figure several 12) highshow electromagnetic several high × energyelectromagneticArea amplitude E was about energy events 24 amplitude× labelled 6 m. The 20, timeevents 21, slice 22 andlabelled analysis 23, respectively.20 (Figure, 21, 22 13) and Theshow 23, shape, severalrespectively. the high dimensions electromagnetic The shape, and thethe energyalignmentdimensions amplitude with and the theevents other alignment excavatedlabelled with 24, tombs 25the and allowother 26, one exrespectively.cavated to interpret tombs The these allowshape, as probable onethe todimensions tombs.interpret andthese the as alignmentprobable tombs.with the other excavated tombs allow one to interpret these as probable tombs.

4.2.5. Area E Area E was about 24 × 6 m. The time slice analysis (Figure 13) show several high electromagnetic energy amplitude events labelled 24, 25 and 26, respectively. The shape, the dimensions and the alignment with the other excavated tombs allow one to interpret these as probable tombs.

Sensors 2019, 19, 3494 14 of 26 Sensors 2019, 19, x FOR PEER REVIEW 14 of 26

SensorsSensors 2019 2019, ,19 19, ,x x FOR FOR PEER PEER REVIEW REVIEW 1414 ofof 2626

FigureFigure 12. 12.Time Time slices slices ofof AreaArea D:D: thethe mostmost signiﬁcantsignificant anomalies, probably probably related related to totombs, tombs, are are highlightedhighlighted at at depth depth of of 28–74 28–74 cm cm (20, (20, 21,21, 22,22, 23).23). A linear anomaly anomaly (19) (19) could could be be related related to toa buried a buried structurestructure or or to to the the boundary boundary of of old old excavation. excavation.

4.2.5. Area E AreaFigureFigure E was 12.12. Time aboutTime slicesslices 24 ofof6 m.AreaArea The D:D: timethethe mostmost slice significansignifican analysist t(Figure anomalies,anomalies, 13 ) probablyprobably show several relatedrelated high toto tombs,tombs, electromagnetic areare highlighted at depth of× 28–74 cm (20, 21, 22, 23). A linear anomaly (19) could be related to a buried energyhighlighted amplitude at events depth of labelled 28–74 cm 24, (20, 25 21, and 22, 23). 26, A respectively. linear anomaly The (19) shape,could be the related dimensions to a buried and the structure or to the boundary of old excavation. alignmentstructure with or the to otherthe boundary excavated of old tombs excavation. allow one to interpret these as probable tombs.

Figure 13. Time slices of Area E: the most significant anomalies, probably related to tombs, are highlighted at depth of 68–87 cm (24, 25) and 111–130 cm (26).

4.2.6. Area F

Area F was about 31.5 × 9 m. The time slice analysis (Figure 14) show several high FigureFigureFigure 13. 13.13. Time TimeTime slices slicesslices ofofof AreaArea E:E: thethe most most significan signiﬁcantsignificant t anomalies,anomalies, probablyprobably probably relatedrelated related toto totombs,tombs, tombs, areare are electromagnetic energy amplitude events labelled 27, 28, 29 and 30, respectively. The shape, the highlightedhighlightedhighlighted at at at depth depth depth of of of 68–8768–87 68–87 cmcm cm (24,(24, (24, 25) 25) and and 111–130 111–130111–130 cm cm (26). (26). dimensions and the alignment with the other excavated tombs allows these to be interpreted as 4.2.6.probable Area Ftombs. The anomaly labelled 27 was attributed to a modern pipeline. 4.2.6.4.2.6. Area Area F F AreaArea F wasF was about about 31.5 31.59 m. × The9 m. time The slice time analysis slice (Figureanalysis 14 )(Figure show several 14) show high several electromagnetic high Area F was about ×31.5 × 9 m. The time slice analysis (Figure 14) show several high energyelectromagneticelectromagnetic amplitude energyenergy events amplitudeamplitude labelled 27, eventsevents 28, 29 labelledlabelled and 30, 2727,respectively. , 28,28, 2929 andand 30,30, The respectively.respectively. shape, the TheThe dimensions shape,shape, thethe and thedimensionsdimensions alignment andand with thethe the alignmentalignment other excavated withwith thethe other tombsother excaexca allowsvatedvated these tombstombs to allows beallows interpreted thesethese toto be asbe interpreted probableinterpreted tombs. asas Theprobableprobable anomaly tombs. tombs. labelled The The anomaly 27anomaly was attributed labelled labelled 27 27 to was was a modern attributed attributed pipeline. to to a a modern modern pipeline. pipeline.

Figure 14. Time slices of Area F: the most significant anomalies, probably related to tombs, are highlighted at depth of 96–142 cm (28, 29, 30). An evident and wavy anomaly at 41–74 cm (27) is identifiable as a modern pipe.

4.2.7. Area G Figure 14. Time slices of Area F: the most signiﬁcant anomalies, probably related to tombs, are FigureFigure 14.14. TimeTime slicesslices ofof AreaArea F:F: thethe mostmost significantsignificant anomalies,anomalies, probablyprobably relatedrelated toto tombs,tombs, areare highlighted at depth of 96–142 cm (28, 29, 30). An evident and wavy anomaly at 41–74 cm (27) is highlightedhighlighted at at depth depth of of 96–142 96–142 cm cm (28, (28, 29, 29, 30). 30). An An evident evident and and wavy wavy anomaly anomaly at at 41–74 41–74 cm cm (27) (27) is is identiﬁableidentifiable as as a a modern modern pipe. pipe. identifiable as a modern pipe.

4.2.7.4.2.7. Area Area G G

Sensors 2019, 19, 3494 15 of 26

Sensors 2019, 19, x FOR PEER REVIEW 15 of 26

4.2.7. AreaArea G G was about 23 × 6 m. In this area an electromagnetic energy penetration depth of 40–45 ns Areawas Gfound. was Figure about 15 23 shows6 m. the In processed this area radarg an electromagneticram related to the energy 1st profile penetration and to the depth 600 MHz of 40–45 ns antenna. It shows several× hyperbolic shaped reflection events. The first reflection event (dashed was found. Figure 15 shows the processed radargram related to the 1st profile and to the 600 MHz yellow)Sensors 2019 continuous, 19, x FOR PEER and REVIEW slightly undulating reflector appears strong and irregular and reaches15 of 26a antenna.maximum It shows depth several below hyperbolic the ground shaped surface reflectionranging from events. 0.8 to The1.30 firstm. This reflection event could event be (dashedrelated to yellow) continuousthe shallowArea and G slightlybedrock was about visible undulating 23 × 6also m. inIn reflectorthethis old area excavati an appears electromagneticons (Figure strong 6). andenergy The irregular reflection penetration andevents depth reaches labelled of 40–45 a 31 maximum nsis depthvisible belowwas found. at the depth groundFigure from 15 surfaceabout shows 0.40 the ranging m processed to 0.60 from m. radarg The 0.8 todimensionram 1.30 related m. (about This to the event 0.80 1st m) profile could and shapeand be related to suggest the 600 to that theMHz it shallow bedrockisantenna. probably visible It alsorelatedshows in toseveral the the old presence hyperbolic excavations of a shapedtomb. (Figure Partic reflecularly6tion). The events.it is reflection important The first eventsto underlinereflection labelled thatevent this 31 (dashed isevent visible at depthshow fromyellow) a about polarity continuous 0.40 inversion m and to 0.60 slightlyof the m. electromagnetic Theundulating dimension reflector wave (about andappears 0.80this is m)strong related and and shape to the irregular presence suggest and of that voidreaches it is[38]. probably a maximumThe time depth slice below analysis the ground(Figure surf16) showace ranging severa froml high 0.8 electromagnetic to 1.30 m. This energy event could amplitude be related events to related to the presence of a tomb. Particularly it is important to underline that this event show a labelledthe shallow 31, 32,bedrock 33 and visible 34, respectively. also in the oldThe excavati shape, theons dimensions(Figure 6). The and reflection the alignment events with labelled the other 31 is polarityexcavatedvisible inversion at depth tombs of from theallow electromagneticabout these 0.40 to be m interpretedto 0.60 wave m. The as and probabledimension this is tombs. related (about 0.80 to the m) presenceand shape ofsuggest void [that38]. it is probably related to the presence of a tomb. Particularly it is important to underline that this event show a polarity inversion of the electromagnetic wave and this is related to the presence of void [38]. The time slice analysis (Figure 16) show several high electromagnetic energy amplitude events labelled 31, 32, 33 and 34, respectively. The shape, the dimensions and the alignment with the other excavated tombs allow these to be interpreted as probable tombs.

FigureFigure 15. Area 15. Area G: reﬂection G: reflection events events probably probably related to to a atomb tomb (31), (31), and and to the to bedrock, the bedrock, at a depth at a depth betweenbetween 0.8 to 0.8 1.30 to 1.30 m. m.

The time slice analysis (Figure 16) show several high electromagnetic energy amplitude events labelled 31, 32, 33 and 34, respectively. The shape, the dimensions and the alignment with the other Figure 15. Area G: reflection events probably related to a tomb (31), and to the bedrock, at a depth excavated tombsbetween allow0.8 to 1.30 these m. to be interpreted as probable tombs.

Figure 16. Time slices of Area G: the most significant anomalies, probably related to tombs, are highlighted at depth of 41–60 cm (31, 32), 69–102 (33) and 110–143 (34).

FigureFigure 16. Time 16. Time slices slices of Areaof Area G: G: the the most most signiﬁcantsignificant anomalies, anomalies, probably probably related related to tombs, to tombs, are are highlighted at depth of 41–60 cm (31, 32), 69–102 (33) and 110–143 (34). highlighted at depth of 41–60 cm (31, 32), 69–102 (33) and 110–143 (34).

Sensors 2019, 19, 3494 16 of 26

Sensors 2019, 19, x FOR PEER REVIEW 16 of 26 4.3. Integration of Topographical and Geophysical 3D Data 4.3.The Integration integrated of analysisTopographical between and Geophysical the 3D topographical 3D Data and geophysical data allowed for a better understandingThe integrated of the analysis nature between of the detected the 3D topographical GPR anomalies. and geophysical In some data cases, allowed the for archaeological a better interpretationunderstanding of GPRof the measurements nature of the detected was favoured GPR anomalies. by displaying In some thecases, acquired the archaeological data set with iso-amplitudeinterpretation surfaces, of GPR using measurements a certain percentage was favoured of the by maximumdisplaying complexthe acquired trace data amplitude set with thresholdiso- valueamplitude (Figure 17 surfaces,). Obviously, using a lowering certain percentage the threshold of the value maximum increases complex the visibilitytrace amplitude of the mainthreshold anomaly value (Figure 17). Obviously, lowering the threshold value increases the visibility of the main and smaller objects, but also the heterogeneity noise. Relatively strong continuous reﬂections are anomaly and smaller objects, but also the heterogeneity noise. Relatively strong continuous visible on the threshold volumes. This visualization technique portrayed better the evidence of the reflections are visible on the threshold volumes. This visualization technique portrayed better the anomaliesevidence found of the in anomalies the surveyed found areas. in the surveyed areas.

FigureFigure 17. Areas17. Areas A-G: A-G: the the collected collected data data displayeddisplayed set set with with iso-amplitude iso-amplitude surfaces, surfaces, using using a certain a certain percentagepercentage of the of the maximum maximum complex complex trace trace amplitudeamplitude threshold threshold value. value.

In thisIn this work, work, an an integration integration of of 3D 3D topographicaltopographical information information with with the the GPR GPR data data results results was was necessary first of all to understand the context and continuity of the structures found. It demonstrates necessary first of all to understand the context and continuity of the structures found. It demonstrates that geophysical analysis can be very useful in order to place archaeological structures into a three- thatdimensional geophysical framework. analysis canArchaeological be very usefuland topographical in order to information place archaeological merged with GPR structures mapping into a three-dimensionalof the buried units framework. leads directly Archaeological to a better unders andtanding topographical of the topography information of the necropolis merged (Figure with GPR mapping18) and of thefacilitates buried the units reading leads and directly therefore to a betterthe interpretation understanding of the of thespatial topography relations between of the necropolis the (Figurevarious 18) and anomalies facilitates at different the reading depths. and therefore the interpretation of the spatial relations between the variousAn anomaliesimportant aspect at diff oferent the depths.research carried out was to identify those methodologies that could illustrateAn important the results aspect of geophysical of the research surveys carried filling outthe communication was to identify gap those that methodologiesoften makes them that hard could illustrateto understand. the results In of this geophysical project the surveys aid of computer filling the graphics communication has been gap necessary that often to realize makes synthesis them hard to understand.images calculated In this project starting the from aid ofthree-dimensional computer graphics environments. has been necessaryThe first step to realizewas to synthesiscreate three- images dimensional models from the georadar iso-surfaces volumes file: the meshes were then correctly calculated starting from three-dimensional environments. The first step was to create three-dimensional oriented and scaled in relation to the size of the individual areas investigated. The polygonal surfaces modelshave from been themapped georadar using a iso-surfaces gradient texture volumes that matche file:s the the meshesaltimetric were map, on then the correctly axis of height, oriented with and scaleda range in relation of colours to the that size from of the mo individualst superficial areas areas investigated. (red colour), The go with polygonal the different surfaces shades have to been mappedindicate using the a different gradient dimensions texture that up matches to that deep the altimetricer (blue colour). map, onIn thethis axisway, of three-dimensional height, with a range of colourssimulations that fromhave been the mostcarried superficial out on the areasperspectiv (rede colour),views of gothe withindividual the di areasfferent explored, shades making to indicate the diimmediatelyfferent dimensions comprehensible up to the that data deeper obtained, (blue both colour). as in the In whole this way,and in three-dimensional the specific levels. The simulations most haveimportant been carried aspect out obtained on the perspectiveis an overview views that ofcan the simultaneously individual areas compare explored, the uncovered making funerary immediately comprehensiblestructures with the those data detected obtained, by boththe geometrical as in the wholedata: a andrelationship in the specificbetweenlevels. the visible The and most theimportant non- aspectvisible obtained able to is provide an overview new informat that canion simultaneouslyon the structures compareand the extension the uncovered of the necropolis. funerary This structures is useful for the development of the most up-to-date considerations of a topographical nature of the site with those detected by the geometrical data: a relationship between the visible and the non-visible able and to offer further elements in the planning of next archaeological excavation (Figures 18 and 19). to provide new information on the structures and the extension of the necropolis. This is useful for the development of the most up-to-date considerations of a topographical nature of the site and to offer further elements in the planning of next archaeological excavation (Figures 18 and 19). Sensors 2019, 19, 3494 17 of 26

Sensors 2019, 19, x FOR PEER REVIEW 17 of 26

Figure 18. 3D reproduction of the anomalies under the soil in the area A inserted in a digital photo: Figure 18. 3D reproduction of the anomalies under the soil in the area A inserted in a digital photo: (A) general view of the anomalies; (B) slice of the anomalies at depth of about 60 cm; (C) slice of the (A) generalanomalies view at depth of the of anomalies; about 90 cm; (B (D) slice) slice of of thethe anomalies at at depth depth of ofabout about 180 60cm. cm; (C) slice of the anomalies at depth of about 90 cm; (D) slice of the anomalies at depth of about 180 cm.

Sensors 2019, 19, 3494 18 of 26 Sensors 2019, 19, x FOR PEER REVIEW 18 of 26

Figure 19. 3D3D reproduction reproduction of the the anomalies under the soil in the area G.

4.4. Archaeological Interpretation The seven areas investigated by georadar prospecting are close to the sector of the funerary area excavated in 1981–1982 and still partially visible (see Section 3.2), lying to the north, south and east of it (Figure7). In the eastern sector, some areas partially overlap the areas investigated during the archaeological campaigns of 1982 and 1985. Considering the maximum depth to which anomalies with archaeological interest are found (m 1.4–1.5 max.), the best results are offered by the 600 MHz antenna, which allows for a detailed reading at different depths. Moreover, these results are always confirmed by the 200 MHz antenna. The time slices of each area were georeferenced in the new topographical map of the necropolis, in order to define a precise location of the anomalies and to compare them with the already known archaeological features. The surface of the investigated areas is flat and the land on which the necropolis extends is slightly sloping from east to west. Geophysical surveys have shown that the rocky bank is not very deep (generally between 0.8 and 1.4 m). It is less deep in the eastern portion of the archaeological park, while it tends to deepen towards north and north-west. Overall, in each investigated area, anomalies of archaeological interest were found, very useful for the study of the organisation and extension of the necropolis. The interpretation of mostly of these anomalies as tombs was facilitated by their georeferencing in the new archaeological map of the necropolis, which collect all the already known archaeologicalFigure 20. Archaeological data. Indeed, themap interpretation of the Monte D’Elia was mostly necropolis based with on the the georeferenced geophysical time-slices characteristics of of the anomaliesthe seven areas and investigated: their sizes and A: ti orientationme-slice at 34–58 in relation cm depth; to the B: time-s alreadylice known at 33–56 tombs. cm depth; C: time- sliceIn particular, at 28–47 cm the depth; georadar D: time-slice survey at has 68–87 highlighted cm depth; aboutE: time-slice 25 anomalies at 41–60 cm that depth; can be F: interpretedtime-slice as furtherat 41–60 tombs cm (chests depth; ofG: slabstime-slice or sarcophagi), at 41–60 cm depth. evidencing the extension of the necropolis beyond the limits known to date. Indeed, further possible tombs were found in a strip of land (Area A) located just 4.4.north Archaeological of the stratigraphic Interpretation excavations carried out in 1982 and in the area located immediately south (Area G) of the large area investigated in 1981; at this regard, the extension of the necropolis in this The seven areas investigated by georadar prospecting are close to the sector of the funerary area direction was already suggested by the isolated tomb found about 25 m south of the 1981 excavation excavated in 1981–1982 and still partially visible (see Section 3.2), lying to the north, south and east area, in correspondence of the enclosure wall of the archaeological park. of it (Figure 7). In the eastern sector, some areas partially overlap the areas investigated during the The results of the B and C Areas, partly overlapped on the sectors excavated in 1981–1982, now archaeological campaigns of 1982 and 1985. Considering the maximum depth to which anomalies filled in, attest the presence of further tombs outside those sectors. In the same way, the results of the D with archaeological interest are found (m 1.4–1.5 max.), the best results are offered by the 600 MHz and E Areas, partly overlapped on the sector excavated in 1985, also filled in, suggest an extension of antenna, which allows for a detailed reading at different depths. Moreover, these results are always the necropolis beyond the boundaries established in 1980s. The latter data seems to be also documented confirmed by the 200 MHz antenna. The time slices of each area were georeferenced in the new topographical map of the necropolis, in order to define a precise location of the anomalies and to compare them with the already known archaeological features.

Sensors 2019, 19, x FOR PEER REVIEW 18 of 26 Sensors 2019, 19, 3494 19 of 26 by the results of the georadar survey of Area F, where other anomalies were found, perhaps also related to tombs; these data are very interesting, since the archaeological excavations conducted in 2004 immediately further south, in correspondence with the building at the entrance of the Archaeological Park, had not found ancient evidence [25]. Obviously, only future stratigraphic excavations could confirm the identification of these anomalies with tombs, but surely the analisys of the results of the georadar surveys will be able to address future archaeological research. Moreover, stratigraphic excavations could allow to identify also further Archaic tombs consisting of small pit graves generally covered by a slab, which are more difficult to detect among the other anomalies highlighted by GPR prospecting in this area. The most significant anomalies of each area are examined below.

4.4.1. Areas A-C Area A has an irregular shape of about 380 m2 and is located immediately north of the area investigated in 1982, overlapping partially its south-eastern end. At the depth of 34–58 cm some anomalies referable to filled cavities (nos. 1–2, 5–6) can be seen (Figure 20). They could be identified with tombs made of chests of slabs or sarcophagi due to their shape and size. In all cases, these anomalies do not seem deeper than 70–90 cm, a depth where the rocky bank already is attested in the southern part of Area A. The rocky bank isFigure sloping 19. towards 3D reproduction north and of north-west, the anomalies where under it the is m soil 1.35 in the deep. area G.

Figure 20. Archaeological mapmap ofof the the Monte Monte D’Elia D’Elia necropolis necropolis with with the the georeferenced georeferenced time-slices time-slices of the of seventhe seven areas areas investigated: investigated: A: time-slice A: time-slice at 34–58 at 34–58 cm depth;cm depth; B: time-slice B: time-s atlice 33–56 at 33–56 cmdepth; cm depth; C: time-slice C: time- atslice 28–47 at 28–47 cm depth; cm depth; D: time-slice D: time-slice at 68–87 at 68–87 cm depth;cm dept E:h; time-slice E: time-slice at 41–60at 41–60 cm cm depth; depth; F: time-sliceF: time-slice at 41–60at 41–60 cm cm depth; depth; G: G: time-slice time-slice at 41–60at 41–60 cm cm depth. depth.

4.4. ArchaeologicalThe identification Interpretation with tombs seems likely for anomalies nos. 5–6, confirming continuation to the north-west of the tombs brought to light in 1982. The most problematic interpretation concerns The seven areas investigated by georadar prospecting are close to the sector of the funerary area the anomalies nos. 1–2; the first falls in an area partially excavated in 1982 and may be due to less excavated in 1981–1982 and still partially visible (see Section 3.2), lying to the north, south and east compact soil following the archaeological excavations, while the no. 2 has a divergent orientation of it (Figure 7). In the eastern sector, some areas partially overlap the areas investigated during the from the nearby tombs. It is possible that it is in some connection with the evident anomaly no. 4, archaeological campaigns of 1982 and 1985. Considering the maximum depth to which anomalies visible from about 50 to 100 cm deep, oriented roughly in the east-west direction and crossing the with archaeological interest are found (m 1.4–1.5 max.), the best results are offered by the 600 MHz entire southern portion of the investigated area (Figure 21). This anomaly perhaps corresponds to a antenna, which allows for a detailed reading at different depths. Moreover, these results are always pipeline, probably modern, which seems to have been laid avoiding the structures that determined the confirmed by the 200 MHz antenna. The time slices of each area were georeferenced in the new anomalies nos. 5–6. There are some modern structures, now largely destroyed, in the north-western topographical map of the necropolis, in order to define a precise location of the anomalies and to sector of the park and it is possible that the pipeline was directed to this area. Between the depths of compare them with the already known archaeological features.

Sensors 2019, 19, x FOR PEER REVIEW 20 of 26 size with a chest of slabs placed very close to the tomb no. 2/82 brought to light in 1982 (see Section 3.2), arranged in a recess of the rocky bank and perhaps documented by another anomaly visible approximately at the same depth. Area B has a rectangular shape of 6 × 30 m (180 m2) and is between the area excavated in 1981– 1982 and the area investigated in 1985. The georadar measurements documented the bedrock already at a depth of 90–100 cm in the central and southern part of the area, while to the north it is instead found at −130/−140 cm. In the layer of soil above the bedrock, various anomalies with archaeological interest have been found. In particular, between 30 and 70 cm of depth we can see two anomalies (nos. 11–12) oriented roughly in the east-west direction (Figures 20 and 23) and referable to cavities, which for orientation and dimension could be identified with tombs, made of stone slabs or sarcophagi, like the sarcophagus no. 5 unearthed in 1981 immediately further north; it was recovered a few centimeters of depth together with a cippus and then transported to the Museo Civico Messapico of Alezio (see Section 3.2). The anomaly no. 10 (Figure 23), visible up to a depth of 90 cm, could have been produced by remains of the excavation activities carried out to bring to light this sarcophagus: the high reflection could be due to the uncompact soil and stones used for filling the archaeologicalSensors 2019, 19, 3494 excavation. 20 of 26 Linear anomalies nos. 13 and 14 (Figure 23), visible in the southern portion of the area between 40 and 70 cm deep, are very interesting. They are arranged to form a right angle and can be referred to1.40 two and walls 1.80 of m, uncertain another anomaly chronology (no. (related 3) can also to a besquare seen (Figurestructure?) 22), setting referable on to the a cavityrocky elongatedbank. In the in northernthe north-west portion/south-east of the investigated direction, area, which between could be85 identiﬁedand 140 cm by deep, shape, there orientation are two anomalies and size with (nos. a 8–9)chest oriented of slabs respectively placed very closein an toeast-west the tomb and no. nort 2/82h-south brought direction to light in(Figure 1982 (see 24): Section they are 3.2 referable), arranged to cavities,in a recess which of the according rocky bank to their and perhapsshape and documented size can be by identified another anomalywith other visible tombs approximately made of stone at slabsthe same or sarcophagi. depth.

FigureFigure 21. 21. EasternEastern sector sector of of the the investigated investigated area. area. A: A:time time-slice-slice at 70–93 at 70–93 cm cmdepth; depth; C: time-slice C: time-slice at 41– at 6041–60 cm depth; cm depth; G: time-slice G: time-slice at 83–102 at 83–102 cm cmdepth. depth.

Area B has a rectangular shape of 6 30 m (180 m2) and is between the area excavated in 1981–1982 × and the area investigated in 1985. The georadar measurements documented the bedrock already at a depth of 90–100 cm in the central and southern part of the area, while to the north it is instead found at 130/ 140 cm. In the layer of soil above the bedrock, various anomalies with archaeological interest − − have been found. In particular, between 30 and 70 cm of depth we can see two anomalies (nos. 11–12) oriented roughly in the east-west direction (Figures 20 and 23) and referable to cavities, which for orientation and dimension could be identified with tombs, made of stone slabs or sarcophagi, like the sarcophagus no. 5 unearthed in 1981 immediately further north; it was recovered a few centimeters of depth together with a cippus and then transported to the Museo Civico Messapico of Alezio (see Section 3.2). The anomaly no. 10 (Figure 23), visible up to a depth of 90 cm, could have been produced by remains of the excavation activities carried out to bring to light this sarcophagus: the high reflection could be due to the uncompact soil and stones used for filling the archaeological excavation. Sensors 2019, 19, 3494 21 of 26

Sensors 2019, 19, x FOR PEER REVIEW 21 of 26

FigureFigure 22. Eastern 22. Eastern sector sector of of the the investigated investigated area.area. A: time-slice time-slice at at156–180 156–180 cm cmdepth; depth; C: time-slice C: time-slice at at 98–116 cm depth; G: time-slice at 110–129 cm depth. 98–116 cmFigure depth; 22. Eastern G: time-slice sector of atthe 110–129 investigated cm depth.area. A: time-slice at 156–180 cm depth; C: time-slice at 98–116 cm depth; G: time-slice at 110–129 cm depth.

Figure 23. Western sector of the investigated area. B: time-slice at 50–73 cm depth; D: time-slice at 68–87 cm depth; E: time-slice at 111–130 cm depth; F: time-slice at 110–129 cm depth. Sensors 2019, 19, 3494 22 of 26

Linear anomalies nos. 13 and 14 (Figure 23), visible in the southern portion of the area between 40 and 70 cm deep, are very interesting. They are arranged to form a right angle and can be referred to two walls of uncertain chronology (related to a square structure?) setting on the rocky bank. In the northern portion of the investigated area, between 85 and 140 cm deep, there are two anomalies Sensors 2019, 19, x FOR PEER REVIEW 22 of 26 (nos. 8–9) oriented respectively in an east-west and north-south direction (Figure 24): they are referable to cavities, which according to their shape and size can be identiﬁed with other tombs made of stone Figure 23. Western sector of the investigated area. B: time-slice at 50–73 cm depth; D: time-slice at 68– slabs or sarcophagi. 87 cm depth; E: time-slice at 111–130 cm depth; F: time-slice at 110–129 cm depth.

Figure 24. WesternWestern sector of the investigated area. B: B: time-slice at 101–124 cm depth; D: time-slice at 68–87 cm depth; E: time-slice at 111–130 cm de depth;pth; F: time-slice at 123–142 cm depth.

Area C C (m (m 11 11 × 14.50;14.50; 159.50 159.50 m m2) 2is) isimmediately immediately west west of ofArea Area B and B and it is it partly is partly superimposed superimposed on × theon thearea area investigated investigated in 1982 in and 1982 now and buried nowburied (see Section (see Section3.2). Between 3.2). Between15 and 60 15cm and deep 60 there cm deepis an anomalythere is an (no. anomaly 18) elongated (no. 18) in elongated a north-west/south-east in a north-west direction/south-east and direction referable and to referablea cavity (Figures to a cavity 20 and(Figures 21). It20 is and located 21). Iton is the located same on alignment the same of alignment the tombs of themade tombs of chests made of of stone chests slabs of stone nos. slabs 1–4 and nos. 6,1–4 discovered and 6, discovered in 1981, inand 1981, could and be could identified be identiﬁed with another with another tomb of tomb the ofsame the samegroup group (and type?). (and type?). The sameThe same interpretation interpretation could could be proposed be proposed for foranomal anomalyy no. no. 17, 17, stretched stretched in in a anorth-south north-south direction, direction, visible a little further to the east between 40 and and 70 cm cm deep deep (Figures (Figures 2020 andand 21).21). Moreover, Moreover, two two large large anomalies (nos. 15–16) 15–16) are are visible visible in in the the northern northern po portionrtion of the the area area (Figure (Figure 22):22): they they are are documented documented between 70 70 and and 130 130 cm cm deep, deep, up up to to the the bedrock bedrock (which (which in in the the investigated investigated area area is already is already attested attested at –110at –110 cm) cm) and and correspond correspond to totwo two sectors sectors of of the the 1982 1982 excavation; excavation; in in particular, particular, the the anomaly no. 15 corresponds in part to the tomb no. 5/82 5/82 (see (see Sect Sectionion 3.2),3.2), and and partly partly maybe maybe related related to to its its excavation, excavation, and perhaps to a second tomb very close,close, just outside the excavated area.

4.4.2. Areas D-G Area D corresponds to a rectangular area of 12.5 × 19.5 m (244 sqm) located east of Area B and partly superimposed (with the north-west corner) to a small portion of the area investigated by archaeological excavations during 1985 and now buried (see Section 3.2). The georadar measurements showed that in the central and southern sector the bedrock is already at 90–100 cm deep, while towards the north it tends to deepen up to 140–150 cm. No anomalies referable to the area excavated in 1985 are noted, with the exception of a linear track oriented in a north-south direction (no. 19), visible from 15 to 130 cm deep and placed at the limits of the archaeological excavation area (Figures 20 and 23). The most significant anomalies (nos. 20–23), instead, concentrate in the central portion of the area and correspond to cavities documented between 40 and 90 cm deep

Sensors 2019, 19, 3494 23 of 26

4.4.2. Areas D-G Area D corresponds to a rectangular area of 12.5 19.5 m (244 sqm) located east of Area B × and partly superimposed (with the north-west corner) to a small portion of the area investigated by archaeological excavations during 1985 and now buried (see Section 3.2). The georadar measurements showed that in the central and southern sector the bedrock is already at 90–100 cm deep, while towards the north it tends to deepen up to 140–150 cm. No anomalies referable to the area excavated in 1985 are noted,Sensors 2019 with, 19 the, x FOR exception PEER REVIEW of a linear track oriented in a north-south direction (no. 19), visible24 from of 26 15 to 130 cm deep and placed at the limits of the archaeological excavation area (Figures 20 and 23). Thedirection most significantin relation anomaliesto a road crossed (nos. 20–23), the necropolis instead, concentratewith the same in thedirection. central In portion the eastern of the sector, area and on correspondthe other hand, to cavities the graves documented seem to betweenhave a predomin 40 and 90antly cm deep north-south (Figures orientation,20, 23 and 24 with): for some dimensions burials andarranged orientations orthogonally (east-west in an and east-west north-south), direction. coherent In this with case, those it could of the be nearby assumed tombs that brought this sector to light of inthe 1985, necropolis they could was be influenced chests of stoneby another slabs orroad sarcophagi. (maybe the ancient Via Sallentina) that ran nearby Area E is extended m 6 24 (144 m2) and is located north of Area B. Its eastern portion is with a north-south orientation.× Lastly, there are not large (for example 2 × 3 m or more) anomalies superimposedattributable to onchamber the area tombs; excavated at the in current 1985, today state buriedof research, (see Section this result 3.2). Inseems the easternto confirm part the of theprevious area, the hypothesis bedrock on is alreadythe absence attested of this at type a depth of tomb of 60 in cm, the deepening necropolises towards of Aletion/Aletium east and north-east. (D’Elia Between2001, p. 20) 30 andaccording 90 cm ofto depththe archaeological two anomalies evidence, (nos. 24–25) contrary referable to the to situation cavities are documented visible (Figure in other 20), whichMessapian due to settlements, their size could such beas identifiedRudiae, Vaste, with Egnatia chests of and stone Lecce. slabs or sarcophagi, similar to those of the nearIn conclusion, excavated area. it can Also be highlig anomalieshted nos.how 24the and results 25 are of orientedthe topographical in a north-south and geophysical direction as surveys many ofat thesenecropolis tombs. of Some Monte anomalies D’Elia not at betweenonly allowed 90 and us 140 to cmcollect depth new can data be reported, for the knowledge including the of onethis (no.important 26), perhaps funerary related area toand the its two organization, northernmost but tombs also for brought planning to light future in st 1985ratigraphic (Figures excavations. 23 and 24). Indeed,Area on F, the extending one hand, over archaeological an area of 9 excavati31.5 mons (about could 284 focus m2), isthe at anomalies the north-eastern probably end linked of the to × archaeologicalnew tombs in park,order into an constitute area where a verification no archaeological of the geophysical evidence isreported. prospecting This results. area is On crossed the other by a modernhand, these pipeline, excavations as documented can increase by the a very knowledge evident anomalyof the necropolis (no. 27) and visible reconstruct between its 40 use and over 80 cmthe deepcenturies, (Figure allowing 20). Starting for the from recovery 80 cm of of further depth, grav in thee goods southern for a and better eastern chronological part of the definition area is already of the attesteduse of the the funerary bedrock. area Here, and, between more cmgenerally, 90 and in 150creasing deep, threethe knowledge anomalies (nos.of the 28–30)socio-economical are visible (Figuresaspects of24 Aletion/Aletiumand 25), which can during be identified the Messapian with cavities era, such dug as into material the rock culture, and can artisan be interpreted productions, as otherimports tombs from filled the Greek with soil world, and stones.etc.

FigureFigure 25.25.Archaeological Archaeological map map of the of Montethe Monte D’Elia necropolis.D’Elia necropolis. It is possible It is topossible highlight to the highlight distribution the thedistribution anomalies the related anomalies to probably related to new probably tombs new with tombs the georeferenced with the georeferenced with respect with to respect the already to the excavatedalready excavated tombs. tombs.

Funding: This research received no external funding.

Acknowledgments: The research in the necropolis of Monte D’Elia was possible thanks to the Municipality of Alezio, which granted all the necessary permissions. The investigations have been carried out by means the equipment provided by the Laboratory of Geophysics for Archaeology, the Information Technologies Laboratory, and the Laboratory of Ancient Topography, Archaeology and Remote Sensing of the Institute of Archaeological and Monumental Heritage of National Research Council of Italy (Lecce).

Conflicts of Interest: The authors declare no conflict of interest.

Sensors 2019, 19, 3494 24 of 26

Area G has an irregular shape of about 23 6 m (about 126 m2) and is immediately south of the × funerary area unearthed in 1981 (see Section 3.2). The bedrock is attested between 80 and 135 cm deep. Between 40 and 90 cm of depth two anomalies (nos. 31–32) are evident (Figure 20), oriented in a north-south direction and referable to cavities, which could be identiﬁed with chests of stone slabs or sarcophagi. Between 70 and 115 cm of depth we can see another anomaly (no. 33) with the same characteristics, but, in consideration of the dimensions, it could have been produced by two tombs side by side (Figure 21). Finally, another anomaly (no. 34) possibly reliable to a tomb can be seen between 100 and 145 cm deep at the eastern edge of the area (Figure 22). It is partially embedded in the rocky bank. If the interpretation of these anomalies is correct, it would conﬁrm the extension to the south of the funerary area brought to light in 1981. Also the depths to which the burials would be found are coherent with those of the area investigated by archaeological excavations.

5. Conclusions The topographical and geophysical surveys carried out in the area of the Messapian necropolis of Monte D’Elia allowed for the collection of important new data on the extension of the funerary area and its topographical organization. In particular, topographical surveys allowed for the production of a new map of the overall necropolis, where the old plans of the tombs unearthed during the 1981–1985 archaeological excavations were georeferenced. In this new plan, thanks to the comparison with the grave goods studied in the 1980s–1990s and the examination of the types of burials, it was possible to distinguish the tombs of the Archaic period and those of the Classical/Hellenistic age, which partly overlap the oldest ones. The geophysical prospecting allowed to highlight numerous anomalies, many of them with archaeological interest, in part related to the tombs brought to light in 1981–1985 and subsequently buried, in part to new tombs (around 25), unknown and yet to be explored. The identification of the new tombs was possible thanks to the georeferencing of the time slices from GPR investigations in the new archaeological map of the necropolis; moreover, the visualization of these anomalies in the virtual environments of the subsoil, produced thanks to the integration of the 3D topographical and geophysical data, was very useful in the interpretation phase (Figure 25). The georadar measurements allowed us to document the extent of the necropolis beyond the limits that were defined at the time of archaeological excavations, both north and south, and especially to the west. In addition, by integrating the plan of archaeological investigations and the results of geophysical prospecting it is possible to retrieve some data on the articulation of the funerary area, only partly conceivable thanks to the archaeological excavations of 1981–1985. In particular, in the western sector the tombs seem predominantly oriented in a north-west/south-east direction in relation to a road crossed the necropolis with the same direction. In the eastern sector, on the other hand, the graves seem to have a predominantly north-south orientation, with some burials arranged orthogonally in an east-west direction. In this case, it could be assumed that this sector of the necropolis was influenced by another road (maybe the ancient Via Sallentina) that ran nearby with a north-south orientation. Lastly, there are not large (for example 2 3 m or more) anomalies attributable to chamber tombs; at × the current state of research, this result seems to confirm the previous hypothesis on the absence of this type of tomb in the necropolises of Aletion/Aletium (D’Elia 2001, p. 20) according to the archaeological evidence, contrary to the situation documented in other Messapian settlements, such as Rudiae, Vaste, Egnatia and Lecce. In conclusion, it can be highlighted how the results of the topographical and geophysical surveys at necropolis of Monte D’Elia not only allowed us to collect new data for the knowledge of this important funerary area and its organization, but also for planning future stratigraphic excavations. Indeed, on the one hand, archaeological excavations could focus the anomalies probably linked to new tombs in order to constitute a verification of the geophysical prospecting results. On the other hand, these excavations can increase the knowledge of the necropolis and reconstruct its use over the centuries, allowing for the recovery of further grave goods for a better chronological definition of the Sensors 2019, 19, 3494 25 of 26 use of the funerary area and, more generally, increasing the knowledge of the socio-economical aspects of Aletion/Aletium during the Messapian era, such as material culture, artisan productions, imports from the Greek world, etc.

Author Contributions: The authors wrote distinct parts of the article: Section1: G.S.; Section2: I.D.; Section 3.1: I.F., F.G.; Section 3.2: G.S.; Section 4.1 and 4.2: L.D.G., G.L.; Section 4.3: L.D.G., G.L., I.F., F.G.; Section 4.4.1: I.D.; Section 4.4.2: G.S.; Section5: I.D. Funding: This research received no external funding. Acknowledgments: The research in the necropolis of Monte D’Elia was possible thanks to the Municipality of Alezio, which granted all the necessary permissions. The investigations have been carried out by means the equipment provided by the Laboratory of Geophysics for Archaeology, the Information Technologies Laboratory, and the Laboratory of Ancient Topography, Archaeology and Remote Sensing of the Institute of Archaeological and Monumental Heritage of National Research Council of Italy (Lecce). Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Ferrari, I.; Scardozzi, G. Contributo Alla Carta Archeologica di Alezio; Scirocco Editore: Bari, Italy, 2016. 2. Bolognese, S. Rinvenimenti di tombe ad Alezio. In Atti del Convegno dei Comuni Messapi, Peuceti e Dauni, VIII; Scientific Committee: Bari, Italy, 1983; pp. 165–173. 3. Zezza, G.M. Alezio: Continuità di Vita in un Centro Antico del Salento; Arti Grafiche Pugliesi: Martina Franca, Italy, 1991. 4. Lamboley, J.L. Recherches sur les Messapiens IV-II Siècle Avant J.-C; Ecole Française de Rome: Roma, Italy, 1996. 5. D’Elia. Alezio—Archeologia e Storia di un Comune del Salento; D’Elia: Alezio, Italy, 2001. 6. Robinson, E.G.D. Recent Australian and New Zeland field work in the Mediterranean Region. Field survey at Alezio, 2001. Mediterr. Archaeol. 2003, 16, 149–170. 7. De Simone, C.; Marchesini, S. Monumenta Linguae Messapicae, I–II; Reichert: Wiesbaden, Germany, 2002. 8. D’Andria, F. Insediamenti e territorio: l’età storica. In Atti del Convegno di Studi Sulla Magna Grecia, XXX; Scientific Committee: Lecce, Italy, 1991; pp. 393–478. 9. Beraldin, J.A. Integration of laser scanning and close-range photogrammetry. The last decade and beyond. In Proceedings of the XXth ISPRS Congress, Commission VII, Istambul, Turkey, 12–23 July 2004; pp. 972–983. 10. Russo, M.; Remondino, F.; Guidi, G. Principali tecniche e strumenti per il rilievo tridimensionale in ambito archeologico. Archeol. Calc. 2011, 22, 169–198. 11. Gabellone, F.; Ferrari, I.; Giuri, F. Digital restoration using Image-Based 3D models. In Proceedings of the 1st International Conference on Metrology for Archaeology, Benevento, Italy, 22–23 October 2015; pp. 478–482. 12. Ferrari, I.; Giuri, F. Rilievo e restituzione grafica delle aree delle Pozzelle di Pirro e di Apigliano. In Conservare l’acqua. Le pozzelle di Zollino tra Memoria Storica e Indagini Scientifiche; Chiga, A., Durante, P., Giammaruco, S., Eds.; Editrice Salentina: Galatina, Italy, 2015; pp. 63–70. 13. Gabellone, F.; Ferrari, I.; Giuri, F.; Chiffi, M. Image-Based techniques for the virtualization of Egyptian contexts. In Proceedings of the 3rd IMEKO International Conference on Metrology for Archaeology and Cultural Heritage, Lecce, Italy, 23–25 October 2017; pp. 601–606. 14. Ferrari, I.; Quarta, A. The Roman pier of San Cataldo: From archaeological data to 3D reconstruction. VARVirtual Archaeol. Rev. 2019, 10, 28–39. [CrossRef] 15. Gabellone, F.; Ferrari, I.; Giuri, F. A quick method for the texture mapping of meshes acquired by laser scanner. J. Geoinform. FCE CTU 2012, 9, 17–26. [CrossRef] 16. Gabellone, F.; Ferrari, I.; Giannotta, M.T.; Dell’Aglio, A. From museum to original site: 3D environment for the virtual visit of finds re-contextualized in their original provenance. In Proceedings of the 2013 Digital Heritage International Congress, Marseille, France, 8 October–1 November 2013; Volume II, pp. 215–222. 17. Gabellone, F.; Ferrari, I.; Giuri, F. The Palmieri hypogeum in Lecce. From the integrated survey to the dissemination of contents. In Proceedings of the 2015 Digital Heritage International Congress, Granada, Spain, 28 September–2 October 2015; Volume I, pp. 247–254. Sensors 2019, 19, 3494 26 of 26

18. Malfitana, D.; Gabellone, F.; Cacciaguerra, G.; Ferrari, I.; Giuri, F.; Pantellaro, C. Critical reading of surviving structures starting from old studies for new reconstructive proposal of the Roman theatre of Catania. In Proceedings of the 8th International Congress on Archaeology, Computer Graphics, Cultural Heritage and Innovation, Valencia, Spain, 5–7 September 2016; pp. 155–161. 19. De Juliis, E.M. L’attività archeologica in Puglia. In Atti del Convegno di Studi Sulla Magna Grecia, XXI; Scientific Committee: Taranto, Italy, 1982; pp. 293–322. 20. De Juliis, E.M. L’attività archeologica in Puglia. In Atti del Convegno di Studi Sulla Magna Grecia, XXII; Scientific Committee: Taranto, Italy, 1983; pp. 503–530. 21. De Juliis, E.M. L’attività archeologica in Puglia. In Atti del Convegno di Studi Sulla Magna Grecia, XXIII; Scientific Committee: Taranto, Italy, 1984; pp. 421–446. 22. De Juliis, E.M. Un quindicennio di ricerche archeologiche in Puglia: 1970–1984. Parte II: 1978–1984. Taras 1985, 2, 177–227. 23. Andreassi, G. L’attività archeologica in Puglia. In Atti del Convegno di Studi Sulla Magna Grecia, XXV; Scientific Committee: Taranto, Italy, 1986; pp. 371–396. 24. Ciongoli, G.P. Alezio. In Archeologia dei Messapi; D’Andria, F., Ed.; Edipuglia: Bari, Italia, 1990; pp. 197–200. 25. De Santis, V.; Congedo, F. Alezio (Lecce). Monte d’Elia; Raggi. Taras 2010, 1, 179–182. 26. De Pascalis, A.L. Storia dei Rinvenimenti Archeologici di Alezio. Carta Archeologica Dell’area Urbana. MSc Degree Thesis, University of Lecce, Lecce, Italy, 2000–2001. Unpublished. 27. Pagliara, C. Materiali iscritti arcaici del Salento. In Annali della Scuola Normale di Pisa, XIII; Scientific Committee: Pisa, Italy, 1983; Volume 1, pp. 1–99. 28. Lombardo, M. Tombe, necropoli e riti funerari in Messapia: Evidenze e problemi. Studi Antich. 1994, 7, 25–45. 29. Semeraro, G. ν νηυσi: Ceramica Greca e Societa nel Salento Arcaico; Edipuglia: Bari, Italy,1997. 30. Santoro, C. Osservazioni preliminari sui nuovi documenti epigrafici prelatini. Ling. Storia Puglia 1981, 12, 33–80. 31. Santoro, C. Le epigrafi di Alezio e i problemi della lingua messapica. ACCMPD 1983, 8, 59–113. 32. Santoro, C. Nuovi Studi Messapici III, Primo Supplemento; Congedo Editore: Galatina, Italy, 1984. 33. Santoro, C. Alezio. Studi Etruschi 1986, 52, 352–370. 34. Santoro, C. Alezio. Studi Etruschi 1991, 56, 411–417. 35. Uggeri, G. La Viabilità Romana nel Salento; Grafischena: Fasano, Italy, 1983. 36. Conyers, L.B.; Goodman, D. Ground-Penetrating Radar: An. Introduction for Archaeologists; Altamira Press: Walnut Creek, CA, USA, 1997. 37. Persico, R. Introduction on Ground Penetrating Radar: Inverse Scattering and Data Processing; Wiley: New York, NY, USA, 2014. 38. Leucci, G. Non Destructive Testing for Archaeology and Cultural Heritage A Practical Guide and New Perspectives; Springer: Berlin, Germany, 2019. 39. Goodman, D. GPR Slice Version 7.0 Manual. Available online: http://www.gpr-survey.com. 40. Conyers, L.B. Innovative ground-penetrating radar methods for archaeological mapping. Archaeol. Prospect. 2006, 13, 139–141. [CrossRef] 41. Goodman, D.; Steinberg, J.; Damiata, B.; Nishimure, Y.; Schneider, K.; Hiromichi, H.; Hisashi, N. GPR overlay analysis for archaeological prospection. In Proceedings of the 11th International Conference on Ground Penetrating Radar, Columbus, Ohio, 19–22 June 2006. 42. Conyers, L.B. Ground-Penetrating Radar for Archaeology; Altamira Press: Walnut Creek, CA, USA, 2004. 43. Conyers, L.B. Interpreting Ground-Penetrating Radar for Archaeology; Left Coast Press: Walnut Creek, CA, USA, 2012. 44. Conyers, L.B.; Daniels, J.M.; Haws, J.; Benedetti, M. An Upper Palaeolithic landscape analysis of coastal Portugal using ground-penetrating radar. Archaeol. Prospect. 2013, 20, 45–51. [CrossRef] 45. Goodman, D.; Piro, S. GPR Remote Sensing in Archaeology; Geotechnologies and the Environment Series; Springer: Berlin, Germany, 2013; Volume 9.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).