1As Simon Stoddart recently pointed out, the achievements of the Etruscans in the organization of their urban spaces are »one of main contributors to European urbanism«, although they have long been neglected in urbanistic research, overshadowed by Greek and Roman cities[1]. The Vulci Cityscape project aims to improve our understanding of the Etruscan urban phenomenon, our notions of the spatial syntax of the urban layout, its emic and etic conception and the related historical and social circumstances, using the example of Vulci , one of the main Etruscan centres, from its origins to the Middle Ages.

2Etruscan cities have long remained invisible in our vision of the ancient landscape, on the one hand, because of the continuity of occupation of many settlements, and on the other hand, due to certain methodological challenges in approaching urbanism from an Etrusco-Italic perspective: these range from the relative lack of consistent and comprehensive modern archaeological data to difficulties relating to the use of heuristic categories and methods, adopted from the Greek and Roman world. In particular, an ever-increasing research and deeper understanding about early Rome served as blueprint for Etruscan urbanism and trapped it within the typical categories of Roman urban forms and processes[2]. A shift away from such a Helleno- and Romanocentric position in recent years has led to an emphasis on the autochthonous character of urban processes in Etruria[3].

3In an effort to make urban spaces more visible and to understand their structure and inner logic, urban archaeology, in general, has undergone a rather dynamic development, both on a methodological and a theoretical level[4]. Projects that deal with urbanism from a regional perspective, placing urban spaces in greater landscape and global contexts, have increased significantly, partly under the theoretical keywords of globalization or urban ecology[5]. Furthermore, there is a renewed interest in the urban form and development of individual cities, mostly because various and constantly evolving, non-invasive methods, e.g., Ground Penetrating Radar (GPR), LIDAR and multispectral imaging, have allowed entire cities to come into view, providing a comprehensive assessment of their urban macrostructure[6]. Exemplary are the recent publications of GPR prospections in Falerii Novi [7], which covered the entire urban area, but also the prospections in Ocriculum [8], to name just a few. Finally, the detailed interdisciplinary study of small-scale stratigraphies produce information-rich micro-narratives, which can shed light on very different aspects of cities[9]. Urbanism has, therefore, been increasingly conceptualized from a more dynamic, networked perspective that questions ontological boundaries, as Rubina Raja and Søren Sindbæk summarized under the keywords, »Urban Networks and High-Definition Narratives«[10]. Other research focused on a more holistic perspective, centred around the concept of the cityscape, theorizing the physical and literary recursive processes that shape urban spaces and our ever-changing perceptions of them[11]. The Vulci Cityscape project approaches the phenomenon of Etruscan urbanism from such a holistic perspective: the detailed analysis of urban spaces not only aims to better understand the emergence and design of urban environments, but to evaluate the social, religious and political dynamics, recursive processes, their cultural and social figuration, and the various challenges that influence urban development and contribute to its resilience to ever-changing historical events.

4When we speak of Etruscan urbanism, we should first emphasize that the Etruscan city per se never existed[12]. Marzabotto [13] and Spina [14], for example, are among the best studied Etruscan cities, but as they lie in Etruria Padana and have their own specific histories, they are hardly comparable with the processes of urbanization in Southern Etruria[15]. The same holds true for Musarna [16], a settlement founded by Tarquinia in the 4th century B.C. or for smaller settlements like Cetamura del Chianti [17]. The latter is programmatically included in the series Cities and Communities of the Etruscans, to facilitate a dialogue between them and to reflect on the question of what defines an Etruscan city[18]. New research in the large cities of Veii [19], Tarquinia [20] and the Latin city of Gabii [21], where different methods (mainly surveys, geophysics and excavations) have been used in a complementary way, substantially improve our knowledge of urban spaces. Geophysical results in Spina were thoroughly reassessed after excavation[22].

5In particular, the more frequent use of non- or minimally invasive methods allows us to survey urban structures on a large-scale, gaining a comprehensive overview, even if only two-dimensional and without absolute chronological and historical depth[23]. Nevertheless, our idea of Etruscan cities remains incomplete and fragmented, often even static, trapped in a timeless situation; still missing is a vision of a city that is variable, dynamic and fluid[24]. We miss the picture of a multifaceted city in which different functional areas are intertwined, linked to greater archaeological and historical developments, and evolving over the centuries as breathing organs of the urban structure. Therefore, a new interdisciplinary project has been initiated in the city of Vulci, that combines archival research[25], geophysical prospections and targeted archaeological excavations[26]. The project is based at the German universities of Freiburg and Mainz, in cooperation with the Soprintendenza Archeologia, Belle Arti e Paesaggio per la provincia di Viterbo e per l’Etruria meridionale and with the Parco Archeologico di Vulci, and is funded by the Fritz Thyssen Stiftung[27] and the Gerda Henkel Stiftung[28].

6The plateau on which Vulci lies covers an area of at least 91.5 ha (light blue in Fig. 1). If the so-called Pozzatella is also considered, the total area of the city measures around 130 ha[29]. The beginnings of the long process of city formation date back to the Early Iron Age, perhaps even to the end of the Bronze Age[30].

7Vulci offers excellent conditions for large-scale investigations, since it was never built over after its definitive abandonment in the 8th century A.D. Intensive research using various non-invasive methods gives an approximate idea of the structure of the city, which seems to be following the natural morphology of the terrain (see below) but lacks detail, whereby some areas of the plateau could not be addressed at all[31]. It remains open as to whether the entire area was inhabited or whether certain empty and unbuilt areas existed, since aerial photographs did not provide clear results in this regard[32]. Of particular importance is the street layout and how it affects the structure of the city. Finally, we wished to establish whether it is possible to distinguish between different functional areas and the way in which the cityscape is organized and shaped. Evaluating the dialectic between different uses of space and the physical layout will complete the urban plan, and help understand the strategies and performative aspects of the urban design.

8We, therefore, selected the entire area north of the so-called decumanus as a case study. This area is of particular interest for several reasons: first, the urban situation is relatively unknown, and although older aerial photos showed numerous anomalies, they could not be analysed and interpreted coherently[33]. At the same time, the area contains a large number of the archaeological excavations on the city’s territory, such as the tempio grande, the various gates and the domus del criptoportico, which, although not or not fully published, can serve as reference points for different periods relating to the diachronic development of Vulci. Various functional areas are to be expected here or have been postulated based on the aerial photographs[34], but they are not clearly located.

9We surveyed an area of around 22.5 ha with magnetometry (dotted in Fig. 1; light blue in Fig. 2; Fig. 3) and two test areas (red in Fig. 2) using GPR, from September 21st to 26th 2020. All measurements were carried out by a team of geophysicists from Eastern Atlas[35]. In summer 2021 and 2022, excavation campaigns were undertaken at a temple which we discovered during the geophysical prospections (green in Fig. 2), therefore, the results of magnetometry, GPR and archaeological excavation could be compared with one another. In order to allow a better comparability of the data, all results from archival research, geophysical prospections and excavations are incorporated into a Geographical Information System (GIS[36]).

10Following a brief overview of older, non-invasive research into the urban area of Vulci, we will present the methods, methodological challenges and results of our geophysical campaign. Subsequently, we will propose a new city plan for the northern region of Vulci and will briefly discuss the different functional areas identified in the process, as well as the consequences for our image of Etruscan urbanism.

11Before going in medias res, a short methodological premise is needed. Etruscan urbanism still lacks a systematic overall assessment and interpretative models, but also an adequate formulation of its own categories. This is also reflected in the language and the terminology we use to define Etruscan cities and their urban spaces, but also relates to social structures and phenomena[37]. For example, the terms »acropolis« and »forum«, as well as »decumanus« and »cardo« were established for Vulci, using terminology typical of both Greek and/or Roman town planning, to make up for the lack of purely Etruscan terms or coherent independent concepts in Etruscology[38]. As these are not mere analytical categories but carriers of specific concepts, this means, from a postcolonial point of view, that we framed Etruscan cities in Greek/Roman terms before we even knew what an Etruscan city could have looked like. Even if there were Roman phases as in Vulci, we should be careful when using such definitions in regard to the Etruscan phases or its development over time, especially if there is a lack of consensus in the identification of certain areas, for example, the »oriental forum«[39]. Nevertheless, some terms are so entrenched in research, serving as reference points and orientation in the topography of Vulci, that it is difficult to avoid them; to maintain a kind of neutrality, we will use certain terms, such as proper names, specifying that it is the »so-called decumanus«, etc.[40].

12Besides the well-known excavations, Vulci has a rich history of non-invasive investigations. Archaeological excavations, which have been ongoing for almost 200 years, have concentrated mainly on the necropoleis, while large-scale investigations in the urban area have remained the exception. The first excavations in the city were conducted between 1834 and 1837 by the notorious Campanari family[41], but the trenches cannot be located with any certainty[42]. Renato Bartoccini began a more intensive activity in the city, concentrated mainly around the so-called decumanus and the northern region of the city, from the 1950s onwards. The entire length of the so-called decumanus and part of the so-called cardo were unearthed[43]. Of particular importance, however, was the discovery of the late archaic tempio grande[44]. Other buildings in the northern part of the city (such as the domus del criptoportico[45]) near the acropolis and south of the decumanus[46], parts of the city walls and gates[47], as well as productive areas near the river plain[48] were excavated in the subsequent decades.

13Renato Bartoccini’s project in the late 1950s and early 1960s was conceptualized from an interdisciplinary perspective and envisaged not only large-scale excavations but also the inclusion of geophysical prospections, other disciplines like geography and the analysis of aerial photographs, so as to approach the city holistically and, in particular, to produce a new plan[49]. The Fondazione Lerici conducted electrical resistivity surveys between 1955 and 1957[50]; however, the death of Bartoccini in 1963 put an end to these projects for the time being. Only in 1996 were the first promising attempts resumed within the framework of the »scuola cantiere archeologica«[51]. Dario del Bufalo used resistivity, magnetometry and GPR comparatively on several areas south of the so-called decumanus, to assess the potential of the different techniques[52]. The electrical resistivity survey was considered to be the only promising approach and the GPR, in particular, gave poor results[53].

14Over the past few years, we have seen a renewed interest in the application of geophysical prospection techniques, although primarily in smaller areas. Between 2015 and 2018, a team led by Maurizio Forte surveyed an area of ca. 10 ha between the tempio grande and the domus del criptoportico and immediately south of them and of the so-called decumanus with GPR[54]. While magnetometry was mostly inconclusive for archaeological interpretation[55], the GPR worked well and confirmed the pattern of the older aerial photographs[56]. In addition, drone photos and multispectral images of the entire plateau were taken, even though the recently published map of Vulci is based mainly on older aerial photos and the results of the GPR[57]. In 2016, a research group led by Corinna Riva from University College London surveyed part of the plateau near the Ponte Rotto alongside the Fiora River[58]. The team used both magnetometry and GPR, however, the magnetometry produced results, whereas the GPR survey results were inconclusive. Finally, in 2019 and 2020, a team from the University of Gothenburg, together with the British School at Rome (BSR) and the Swedish Institute of Classical Studies in Rome conducted magnetometry and GPR surveys, the former without any tangible results and the latter with particular success, on an area of around 2.1 ha in the eastern part of the city, north of the so-called decumanus and between the edge of the plateau and the so-called cardo, showing some evidence of the Roman occupation in this part of the city[59].

15The Fondazione Lerici was already using aerial photographs in the late 1950s[60], taken between 1938 and 1945, however, only Dinu Adameşteanu developed the first city map from this[61]. Giorgio Pocobelli provided in the late 1990s and early 2000s a detailed map of the city[62] and gave a first glimpse of this »invisible« city, even though not all features discernible in the aerial photographs could be clearly identified, vectorized and mapped. He obtained the best results for the central area of the plateau, especially south of the so-called decumanus, where numerous buildings, some of them public, could be identified[63]. Other areas remain »empty«, due to the aforementioned reasons or because nothing was visible on the aerial photographs; this concerns, for example, both the southernmost sector and the area north of the so-called decumanus. As already mentioned, the new map of the team of the Duke University is based mainly on the same set of aerial photographs as the one of Pocobelli, but can fill some of the blank spots due to the use of new methods in georeferencing older aerial photos and a combination of other non-invasive methods[64].

16Whereas previous geophysical research focused solely on small areas of the city and the aerial photographs provided only partial results, the project presented here considers a larger contiguous area of the city, combining geophysical methods and archaeological excavations in a multiscalar and interdisciplinary approach, to obtain a better understanding of the urban structure and its diachronic development.

17The boundaries for the magnetometer survey are defined to the south by the so-called decumanus and to the west, north and east by the morphology of the plateau or by the fences that surround the archaeological park, defining an area of 22.5 ha (dotted in Fig. 1; light blue in Fig. 2). After preliminary evaluation of the results of the magnetometry (Fig. 3), two smaller areas have been selected for evaluating the operational capability of a GPR: that in the north (GPR1 in Fig. 2) measures 600 m2; the other, near the decumanus, (GPR2 in Fig. 2) 3600 m2.

18In magnetometer surveys, slight changes in the Earth’s magnetic field, originating from objects and structures lying in the ground are detected, recorded and displayed in a 2d-map[65]. As targeted archaeological anomalies from different time periods might be overlayed by magnetic anomalies, caused by geological formation and landscape forming processes, as well as modern impacts, small survey area results are often inconclusive. Therefore, almost complete coverage of the entire area of interest allows a better differentiation of the subsurface structures. This approach is traditionally implemented in Roman archaeology, as shown by the examples of Wroxeter , Carnuntum and Ostia , amongst others[66]. The corresponding technology for large-scale surveys has developed enormously in the recent decade. Mobile, multi-channel devices with RTK-GNSS positioning make it possible to prospect large areas with high resolution. These advantages are implemented in the flexible, mobile, multi-channel system LEA, developed by Eastern Atlas[67]. In Vulci, the LEA MAX system, with 10 Förster FEREX CON 650 gradiometer probes, was used (Fig. 37). The magnetometers measure variations in the vertical difference of the Z-component of the Earth’s magnetic field, with an accuracy of ±0.1 nT. For the positioning, an RTK-GNSS, with a fixed base antenna and a rover antenna, mounted on the LEA cart, was used, resulting in a real time kinematic solution of ±2 cm for the positioning data.

19The GPR method is based on the transmission and reflection of high-frequency pulses of electromagnetic waves in the ground. Objects and interfaces with different electrical properties, especially the dielectric permittivity ε, cause diffractions and reflections of the electromagnetic waves, which allow to record and calculate the position and depth of the objects[68]. The most recent technical developments are moving towards multi-channel array systems, using antennas for inline and crossline measurements, resulting in a very high resolution of shallow archaeological objects and structures. A high resolution with a greater depth can be achieved with multi-channel devices, with relatively low frequencies or arrays and using the full spectrum of frequency[69].

20For the GPR surveys in Vulci, a pulseEKKO Spider system (manufacturer: Sensors & Software Inc., Canada) with four 250-MHz antennas in parallel was used (Fig. 38). However, structures very near to the surface are quasi undetectable, due to the bistatic orientation of the antennas (transmitter and receiver antenna are separated). Again, differential RTK-GNSS, as described above, was used for the positioning of the data. The relatively high electrical conductivity of the pyroclastic soils and rocks in the survey area (see also below) results in a relatively high attenuation of the electromagnetic waves, providing high-resolution images to depths of approximately 1.8 m.

21The interpretative framework is based on the classification of specific magnetic anomalies into three main areas: we distinguish between anomalies originating from modern structures and objects, from natural occurring geological and soil features, as well as anomalies with archaeological evidence. To classify the magnetic anomalies, we define characteristic attributes, such as shape, intensity height, orientation and contextualization within the existing topography, and the nature of the surface.

22The geology of the study area is dominated by the volcanic rocks of the Vulsini Volcanic District[70] (Fig. 4). These include pyroclastic sediments of the Grotte di Castro, Sorano 1 and Sorano 2 formations, as well as volcanic effusive rocks[71]. The pyroclastics form deposits of ash and tuffs, some of which are several metres thick and occur extensively in the survey area. The anthropogenic depressions, which are clearly visible in part in aerial photographs[72], are formed into the tuff of the Sorano 2 formation. Below this formation, thick magmatic rock (tephrite) of up to 35 m occurs.

23Extended anomalies with high intensities and a dipole character are mainly classified as geological disturbances. These can be related to rock and soil features and are often associated with basalt, either bedrock or eroded outcrops. Two extensive areas, one close to the northern gate (green in Fig. 5 [73]) and the other at the western end of the plateau (the so-called acropolis) facing the river (Fig. 6), show these types of geological anomalies. Here, the layer of the Sorano 1 formation is probably eroded, and the high intensities are caused by the underlying tephrite. Within these areas it becomes difficult to recognize archaeological features. Furthermore, distinct, radial-shaped dipole anomalies with very high intensities are visible, caused by lightning-induced remanent magnetization (LIRM)[74], often following the shape of existing structures as lightning-current pathways (Fig. 7). Anomalies of modern origin usually manifest themselves in mostly regular dipole anomalies with a very high intensity (blue in Fig. 8) and are related to visible and known modern structures, i.e. the fences of the archaeological park or the railroad used by the lorries during the old excavations (Fig. 9).

24The archaeological anomalies are divided into eight different types (Fig. 10). Round to oval anomalies with a diameter of around 0.5 m to 2.0 m and a predominantly positive intensity, but without a clearly delineated outline, are interpreted as pit fillings (brown in Fig. 11).

25Round or rectangular negative anomalies indicate the removal of the tuff bedrock for depressions or deepenings like floors or cisterns (coral in Fig. 6 and Fig. 12, see below). Those with a rectangular shape probably served to level the subsoil and constituted the floor of individual rooms. A number of depressions in the central area lie directly next to one another (coral in Fig. 12). The resulting individual rooms vary from 12 m2 to 42 m2, but many may be even smaller, as the individual wall features suggest. Another concentration of rectangular deepenings is located east of the ditch on the so-called acropolis (Fig. 6). In this area lie several anomalies with strong negative intensities; those with a roundish shape are presumably cisterns (Fig. 6). Their diameter varies depending on the dynamics of the dataset. This cluster may be related to the nature of the soil, probably eroded to such an extent that no other layers or sediments remain on the surface, making the depressions particularly visible.

26Anomalies with an extended shape that has hardly any appreciable change in its intensity could be interpreted as trenches or ditches, as in the case of the wide ditch of the inner-urban fortification (the so-called agger), dividing the plateau at its narrowest point and characterized by the lack of any kind of structures (orange in Fig. 13). Since the filling probably consists of local sediment, the ditch is difficult to distinguish from its surroundings, particularly along the southern edge.

27Linear anomalies with high intensities, which are either positive or negative or alternate between positive and negative, define individual walls or wall courses (pink in Fig. 14). We often register a strong bipolarity of individual anomalies, lined up in a row with an average diameter of approx. 0.7 m at a dynamic of ±30 nT. The variance in size of the individual anomalies enable conclusions to be drawn in relation to the thickness of the walls. Less visible, mostly negative linear anomalies can also be assumed to be walls. The polarity of the intensity is caused by the magnetic properties of the building material, which is presumed to be tuff. Linear, sometimes curved anomalies with a low negative intensity[75] are interpreted as stone-built channels or water pipes for the purposes of supply or drainage (light blue in Fig. 14 and Fig. 15). At a dynamic of ±30 nT, such an anomaly measures on average 0.7 m in width.

28Roundish to oval anomalies, with very high positive intensities, are usually caused by thermoremanent magnetization of material in the soil, as this occurs in pit fillings, e.g. with burnt clay as the residue of a furnace or in residues of slag in extraction pits. A series of such anomalies define at least four kilns located at 2.5 to 3.0 m next to one another (red in Fig. 16). On average, the anomalies measure around 4.2 m in length and around 1.7 m in width and are oriented from south-east to north-west. Other anomalies, visible near the north-west »kiln«, also show particularly high intensities. It is likely that this ensemble of anomalies reflects the remains of a workshop area. The immediate surroundings have an almost rectangular shape with a slightly negative intensity, which is likely to point to an enclosed structure.

29Elongated linear anomalies with positive intensity, often accompanied on both sides by narrow, chain-like dipole anomalies with medium intensities (brown in Fig. 17), are attributed to the compaction of the soil close to the surface, which is typical in the case of paths and may be interpreted as road paving, producing different magnetic signatures depending on the material used. The reconstruction of streets is based on different evidence in the magnetometer survey data: sometimes the streets themselves are relatively clear, sometimes they can be reconstructed based on accompanying walls or, in rare cases, channels; in other cases, such as the continuation of the so-called cardo, the data most certainly correspond to a paved street, providing comparisons for the signal of such anomalies.

30An additional category defines unclear anomalies that cannot be unequivocally classified into one of the described types. There are two marked anomalies, both located in the western area, north of the tempio grande (okra in Fig. 18). Both have an elongated shape, are not particularly wide (around 0.6 m), and show a slightly negative intensity. The western trace manifests itself as a series of adjacent, roundish, negative anomalies. The eastern trace varies and is more easily recognizable in the northern region. Interestingly, both do not follow any visible structures, neither buildings nor the existing road system. However, both anomalies run slightly parallel to one another from the north-west to the south-east at a length of 114 m and 160 m, respectively.

31We pursued several goals using the GPR survey: Firstly, to compare the results of GPR and magnetometry in an area where the results of the latter were not always clear. Secondly, we wished to assess the reliability of both methods for detecting streets and wall courses. GPR1 was positioned accordingly in an area in which a road course, flanked by wall structures, was recognized in the magnetometry (Fig. 19, left and central). As the road constitutes the potential northward extension of the so-called cardo, one of the main streets in Vulci, the area is of particular significance.

32As seen in Fig. 20 a–e and Fig. 21 a–e, we obtained the best results up to a depth of 1.0 m. The interferences at the northern edge of the area can be attributed to a modern path. The antique road, however, which is clearly visible in the magnetometer data, appears in the GPR slices only through the accompanying linear structures, presumably walls, on both sides (Fig. 19, above right). These linear structures are best visible at depths between 0.6 and 1.0 m (Fig. 20 d–e and 21 d–e). In contrast, wall courses, which are not recognizable in the magnetic data or show only a very weak magnetic response, are distinctly visible in the GPR data and draw an intricate overlap of structures at different depths (compare Fig. 19 below). However, no coherent buildings or ground plans can be reconstructed. These features are best visible at depths of up to 0.8 m (Fig. 20 a–d and 21 a–d). It appears, therefore, that the road has, at some point, been superseded by higher lying structures.

33It can be concluded that both methods have their own advantages but are best used together, as they complement each other in terms of resolution, dimension and material properties of the archaeological targets. Magnetometry is well suited to identify linear structures, such as roads, while GPR allows to identify (non-magnetic) walls and small-scale features.

34GPR2 was defined to obtain a more detailed picture of the temple identified in the magnetometer data (Fig. 22) and, in particular, to determine the dimension of the walls and to define a promising area for archaeological excavation. Furthermore, as the orientation of the temple deviates from adjacent north-south oriented streets, we included the immediately adjoining areas in the GPR to better understand the local layout and organization of the urban space (Fig. 23). In 2021, we started an excavation at the north-east corner of the temple[76], which now allows for a further methodological comparison and anchors the geophysical prospection results to a stratigraphic sequence.

35Around 20 m north-west of the tempio grande, the magnetometer survey revealed a huge building structure, the plan of which clearly corresponds to a temple: a rectangular cella with pronaos and antae walls is enclosed by three walls to the north, west and east sides; a south wall is not clearly delineated in the data (Fig. 22). The walls of the outer building measure around 40 m by 27 m; the temple is almost north-south oriented and is rotated by 3° to the west.

36The intensities of the magnetic anomalies in the southern part of the area are smaller compared to the other sides. An elongated dipole chain seems to continue along the western external wall of the temple; it is not directly connected with it but probably belongs to a group of similar traces to the west. In the south-east corner of the survey area, a strong dipole anomaly points to another wall of around 11.5 m in length, which is in an exact north-south alignment and is intersected at both ends by further walls.

37The GPR data provide information up to a depth of 1.8 m; up to 0.2 m, no coherent structures are recognizable (Fig. 24 a and 25 a). At 0.2–0.4 m (Fig. 24 b and 25 b), more features with high reflection amplitudes are visible in the picture, however, it is still difficult to recognize archaeological features. The eastern part of the surveyed area is dominated by anomalies, which can be assigned to the modern road. At a depth of 0.4 to 0.6 m (Fig. 24 c and 25 c), parts of a smaller building, probably also a temple or, in general, a public building, can be identified in the south-east. Two parallel walls can be traced for approximately 9.9 m in length. Roundish features indicate the likely position of pillars or columns. At a depth interval of 1.0 to 1.2 m, these three anomalies, with a diameter of 2.0 m, become clearer and indicate a regular distance of ca. 2 m.

38In the very northern part of the surveyed area and north of the newly discovered temple, several features, parallel to one other and to the northern wall of the new temple, appear clearly at a depth of 0.6 to 0.8 m (Fig. 24 d and 25 d) and can be followed to a depth of at least 1.2 to 1.4 m below the actual surface. These can possibly be associated with a street or accompanying walls, respectively, or channels.

39The ground-plan of the temple is distinctly visible at a depth of between 0.6 m and 1.4 m (Fig. 24 d–g and 25 d–g). It is indeed striking that the temple is barely discernible and that the podium walls are distinctly narrower than in the magnetometry. Other structures to the south and most clearly visible at a depth of between 0.6 and 1.2 m (Fig. 24 d–f and 25 d–f) are unlikely to be aligned with the temple: several walls in the south-western part of the survey area probably form part of a built-up area, along with regular buildings north of the so-called decumanus, which stretches to the city walls to the west[77]; these are also related to a street running north-south and clearly visible in the magnetometer survey data. To the east, we wished to highlight, amongst other interesting magnetic anomalies, two parallel features, oriented south-west – north-east, between the cella of the new temple and the smaller public building.

40The first two excavation campaigns in the north-east corner of the new temple in 2021 and 2022 have confirmed the results of the magnetometry and the GPR of a monumental foundation wall, both in the cella and at the peristasis, in varying states of preservation[78] (Fig. 26).

41It was possible to stratigraphically distinguish different areas in our trench: inside the cella and between the foundation wall of the cella and the outer peristasis wall, we found the original, mostly undisturbed filling of the construction layers of the late 6th/early 5th century B.C., thus confirming, for now, the constructive coherence and synchronicity of both walls[79], with the caveat that the excavation between the cella wall and the peristasis wall have not yet been completed. The foundation walls of the cella are extremely well preserved. They were found 0.2 m below the current surface, at an elevation of 72.15 m, and are preserved up to 2.5 m at the deepest point, considerably better than the results of the GPR had suggested. The wall consists of a huge tuff ashlar structure of 1.8 m in width, which is conserved up to five courses high.

42The area of the peristasis wall, by contrast, is exhibited through various spoliation and destruction layers, which, as far as one can determine at present, seem to date to the Early Imperial period. The tuff stones of the upper courses have mostly been robbed, therefore, the coherent sections of the wall only appear between ca. 1 and 1.4 m below the highest point of the cella wall at 71.12 and 70.76 m, respectively. The peristasis wall measures approximately 3 m in width and is covered with a row of runners made from nenfro. The results of the excavation so far indicate that the GPR does not display the full width of the peristasis wall, which holds true in particular for the northern portion, whereby the data of the magnetometry correspond closely with the real width of the walls.

43In the area of the road, we have to distinguish between the eastern and the western half. In the western half, we found a compact horizon of clayey soil, which could cautiously be interpreted either as a walking surface or as a preparation layer. This presumed road was defunctionalized by a north-west – south-east running cut, which was filled in the east with several layers dating back to the late 1st or 2nd century A.D.[80]. Over the entire length of the northern edge of our trench we found a structure of tiles, stones, and mortar, which delineates the road to the north.

44At first glance and contrary to earlier attempts in other areas at Vulci[81], the temple layout appears clearer in the magnetic data than in the GPR time slices. Wall features stand out clearly and the strength and polarity of the anomalies provide important information, e.g. the thickness and material properties of the walls (Fig. 22), which point to the different types of targeting subsurface objects, using the wave propagation method, GPR and potentially magnetic field surveys. In some areas, however, the results complement one another. Whereas the magnetometer survey data show superimposed buildings in the south and south-west of the area (Fig. 22), the GPR results are more detailed (Fig. 23). A group of magnetic dipole anomalies, south of the temple, appear in the magnetic data and are slightly washed out compared to the clearly defined temple walls. Overlapping structures and disturbances, probably caused by more recent construction activity, influenced the signal. Certain other features, such as the internal structure of the smaller cult or public building to the south, remain largely invisible in the magnetic data, which point to the type of stone used, likely to be limestone. In contrast, a relatively detailed ground-plan appears in the GPR data: the internal parallel wall, three roundish pilaster or column foundations made of caementicium – as evidenced by the partly excavated eastern wall – and the northern wall. The different signals in the GPR and magnetometer data could be explained by the use of various building materials with different physical properties, which is substantiated by the ongoing excavation. This confirmed that the magnetometry can very well detect tuff stones, particularly as the foundation walls of both the podium and the cella were encountered at the location indicated (Fig. 26). The GPR provides slightly different structural information: one cannot trace the borders to the north and to the south exactly, while the blocks of the cella and the eastern wall appear in the same position as on the excavation and with approximately the same thickness. Significantly, the foundations in the north show a different picture, maybe because the spoliation layers alter the responsiveness of the tuff blocks (Fig. 26). An amorph anomaly in the north-eastern part of the GPR area, which is discernible at a depth of between 0.4 and 1.0 m (Fig. 24 c–e and 25 c–e) turned out to be a pit of the spoliation of the presumed peristasis wall of the new temple, filled with a considerable number of roof tiles and stones[82]. The structure which accompanies the putative road to the north is also clearly visible in the GPR (Fig. 26). A slight diagonal anomaly, between a depth of 0.6 and 1.0 m (Fig. 24 d–e and 25 d–e), can now be better understood in light of the excavation: it appears to follow the diagonal line of antique cuts through the road and the eastern part of the podium[83]. Moreover, the blocks of the foundation wall of the cella already appear at a depth of 0.2/0.3 m from the present ground level and continue up to a depth of 2.5 m, hence sooner and deeper than the GPR detected them, since the temple appears clearly at a depth of ca. 0.6 m and seems to disappear at a depth of 1.4 m in the data. This will allow for the future calibration of the applied techniques, and highlights the possibilities and limitations of certain geophysical methods under the specific survey conditions in ancient Vulci.

45Overall, we obtained a higher spatial resolution of the building structures and detailed depth information with GPR. Moreover, features detectable by the magnetometry could be identified, which could point to additional buildings (i.e. the rectangular structure in front of the temple in Fig. 23). The archaeological excavation was not only able to date the structures, but also made significant adjustments to the archaeological interpretation of the geophysical data. The potential of the combination of different non-invasive and invasive methods must, therefore, be emphasized.

46The results of the extensive geophysical surveys (Fig. 3) have successfully revealed a whole new map of the northern area of the city (Fig. 27) as a complex palimpsest, in which all time phases can be viewed side by side at the same level. It was possible to recognize different functional areas and to address their reciprocal relationship, thus obtaining a more complete vision of the cityscape of Vulci. In the following, we will briefly discuss the main functional areas that can be identified from the magnetometry.

47The so-called agger that separates the so-called acropolis from the rest of the city and which is visible both in the terrain and in the aerial photographs[84] emerges in the magnetic map as a broad strip, largely free of anomalies and structures, and measuring approx. 20 m at the narrowest point and 40 m at the widest; a slight depression and a relatively steep increase correspond to this anomaly in the terrain (orange in Fig. 13 and Fig. 28). At the crest line of this increase, two parallel linear structures are visible that can most likely be interpreted as walls (pink in Fig. 13 and Fig. 28), but we don’t see here the classical picture of two walls containing an agger. By contrast, the geophysical data of the agger from Gabii show two walls, one sub and the other super aggerem, accompanying the ditch respectively ditches[85] in a different position, leaving our data open for further interpretation. The situation also seems to differ from that at the western gate, where Moretti Sgubini dates a first agger to the middle/second half of the 8th century B.C.[86]. Still, the question remains whether this fortification benefitted from the geological margin or if it is wholly or partially man-made[87]. Acqua Acetosa and its defensive systems of the Final Bronze Age and Early Iron Age seem to be the closest parallel for our situation[88]. Some of the anomalies in the area of the access to the acropolis, which opens into the central artery of the acropolis, can perhaps be interpreted as a monumentalized or at least a densely build-up entrance or maybe as a bastion[89].

48The road network shows the complex, diachronic palimpsest and development of the city structure which highlights the historical depth of Vulci[90]. Different responses of streets in the magnetometer data could reflect various construction techniques and eventually correspond to different phases[91]. The gates and the excavated street sections provide good references for the interpretation, just like the results of other recent geophysical studies[92]. Thanks to the overlapping of buildings and streets at certain crucial points (highlighted in orange in Fig. 29), at least two distinct road systems can be recognized, suggesting different phases in the structure of the city. This, in turn, has far-reaching consequences for our ideas of social organization.

49Streets that seem to belong together and could cautiously be assigned to ›systems‹ are highlighted in blue and red in Fig. 29. This does not preclude the fact that at least individual streets or parts of one system could have been used also by another system. The situation presents itself as being rather complex and we merely wish to provide an initial stimulus for discussion. We clearly distinguish a phase with a relatively regular, orthogonal plan (red in Fig. 29), suggesting a deliberate division of the land and a fundamental (re-)organization of the city layout. The street immediately behind the new temple running east-west[93] is intersected regularly by streets oriented north-south, oscillating in their distance from one another between 55 m in the western area of the plateau and 35 m in the eastern area. Here, the street grid exhibits more regularity or is, at least, more easily legible.

50The streets to the west of the so-called cardo of the second (and presumably) later system (blue in Fig. 29) overlap the major east-west street of the first system. This second system is centred around the so-called cardo with another main north-east – south-west street to its east and a street connecting the tempio grande with the so-called cardo and the northern gates to its west. These three streets are interconnected with secondary collector roads which are, at least in one instance, visible at the excavated section of the cardo (Fig. 30). This system seems to be related to structures that resemble rather large units, organized around various courtyards (e.g. atria, peristyles) and mirroring in some cases the plan of Roman domus, which will be discussed later. The main axes, especially the so-called decumanus and cardo, follow the course of the terrain and connect the natural accesses to the plateau[94]. They do not seem to have changed over time since they match both systems, thus representing the more stable elements of the urban road system, linking the city with the outside (green in Fig. 29).

51Although overlapping roads and structures allow to establish a relative chronology, any attempt to precisely date the road systems without stratigraphic data presents in most cases methodological problems[95].

52To the north of the so-called cardo (Fig. 16. 31) and on the north-western borders of the plateau (Fig. 27), several kilns can clearly be recognized, based on their shape and size[96]: a round main part corresponds to the combustion and firing chamber, a smaller rectangular part to the stoking chamber. The kilns are clustered in certain areas and the four, east of the so-called cardo, are clearly lined up in a row; other kilns are present in their vicinity (Fig. 31). These kilns can be linked with large-scale manufacturing inside the city walls[97]. To date, only one kiln from the Campanari excavation, which can no longer be located[98], was identified in the city area; other productive areas are archaeologically documented in the area of the porta est[99] and ovest[100], north of the city[101], as well as in the Fiora plain[102]. These various known production areas date from the 6th to the 1st century B.C.

53The results of the magnetometry show an intense and dense settlement for the entire area[103], although it cannot be determined whether all buildings existed simultaneously. Only on the so-called acropolis is the picture less clear, due to the impact of geological formation on the magnetic data. It is possible to recognize structures carved into the natural ground[104], often accompanied by walls and clustered in the central part of the area and on the so-called acropolis (Fig. 6. 12. 32). Magnetic anomalies which resemble rows of single stones, mostly following the sides of the streets appear to be connected to these features and could be interpreted as water or drainage channels[105] (light blue in Fig. 14, 15 and 32).

54In other areas of the plateau, chains of dipole anomalies can be addressed as walls[106] (Fig. 33). The rather unclear image, showing positive and negative anomalies around the second, blue road system (Fig. 29) could be associated with the construction technique: collapsed terracotta roofs or walls, either of different building materials with different magnetic properties, e.g. a mixture of basalt, limestone and tuff, as seen in the local opus caementicium (Fig. 30) or the use of bricks, both pointing to a Roman date, which is also confirmed by surface finds and the layout of the buildings, often larger units organized around courtyards. The excavations along the so-called decumanus and cardo, e.g. the domus del criptoportico (2nd/1st century B.C. with renovations in the Julio-Claudian period[107]) and GPR measurements to the west of the domus del criptoportico and between the cardo and the edge of the plateau seem to corroborate this impression. In the latter case, the GPR measurements of the BSR, University of Gothenburg and the Swedish Institute of Classical Studies in Rome showed several buildings with orientations and groundplans consistent with a Roman occupation[108]. Between the tempio grande and the domus del criptoportico Maurizio Forte and his team interpret an impluvium, and to the east, less evident, a peristyle[109]. It could be that the settlement activity was probably concentrated in this area at the time, although not exclusively.

55Some walls overlap one another and the road system, linked to the east-west street, which runs parallel with the so-called decumanus (belonging to the red system; Fig. 29); this street – with due caution – could, therefore, have preceded the building and been obliterated by it.

56In the western area of the plateau (in particular, north-west of the new temple and south-east of the porta nord) or at the highest point of the so-called acropolis, certain anomalies point to small buildings, which are more difficult to categorize (Fig. 34) and which were not visible in aerial photography[110]. Some of them do not seem to fit in a coherent system and are concentrated in different clusters, mostly around elevated positions, thus, they eventually represent a peculiar, fragmented occupation of selected areas of the plateau, in phases not directly connected with one of the two proposed street systems. Others seem oriented to the red street system.

57West of the tempio grande, a new temple, measuring circa 40 m × 27 m[111], as well as another public or sacred building[112] (22 m × 13 m) to its south were discovered (Fig. 22. 23). The three buildings seem to have formed a »sacred district« immediately north of the so-called decumanus (Fig. 22. 23. 35). Both the new temple and the smaller building at the decumanus can be observed in the GPR[113], showing considerable preservation and thus the possibility of encountering intact archaeological stratigraphy. This and the preservation of the temple to a depth of 2.5 m were, indeed, confirmed during the excavations of 2021 and 2022[114].

58Although the newly discovered temple is oriented towards the decumanus, it does not connect to it and has a different orientation from the other sacred buildings. It is relatively striking that all three buildings of the sacred district are oriented differently, and it seems that the two monumental temples, in particular, do not fit well into the surrounding street system. The comparison with Veii or Tarquinia[115], where the monumental structures are likewise not always aligned with one other and/or with the street grid, suggests that the different orientations could be shaped by other ritual or functional needs or that the different directions are dependent on the successive rebuilding phases of the city.

59The basic structure of the new temple can be compared, both in terms of plan and proportions, with some well-known temples from other Etruscan cities and points to the architectural form of Graeco-italic peripteroi. These include the neighbouring tempio grande, reconstructed by Giuseppe Sassatelli and Elisabetta Govi as a tetrastyle peripteros with six columns at the sides and five columns at the back[116], which, according to some architectural terracottas found nearby during its excavation, has a first phase in the late 6th century B.C.[117]. Furthermore, the temple of Tinia in Marzabotto and temple B in Pyrgi are suitable comparisons, although smaller. Both are peripteral buildings: the Tinia temple dates from the early 5th century B.C.[118]; temple B at Pyrgi from the late 6th century B.C.[119]. After the preliminary analysis of the excavation data, the new temple would also date to the end of the 6th or beginning of the 5th century B.C.[120].

60While the new, late archaic temple was spoliated in the Early Imperial Era, the small public or sacred building to its south corroborates with a presumed Augustean phase of the tempio grande[121], a continuation of this sacred quarter in the late Republic and Imperial period. The visible part of the east side of the building shows masonry and pillars of opus caementicium, covered with limestone ashlars, which indicate a late Republican or Imperial date. The west and north wall, as well as the west pillars of this building, stand out very clearly from a geophysical perspective (Fig. 22. 23). This suggests a rather hypothetical reconstruction of a small peripteros sine postico, which suggestively recalls the model of a small temple from the votive deposit of the porta nord – without suggesting an identification –, with angular columns on the façade and half-columns on the walls, dating back to the 2nd century B.C.[122].

61The results of the geophysical surveys not only changed our ideas of the cityscape of Vulci, but also contributed to a better comprehension of Etruscan urbanism in general, particularly with regard to the internal organization of the city and to the significance and layout of sacred spaces. We gained, for the first time, a detailed georeferenced plan of Vulci’s northern area, which differs significantly from former attempts[123] and which allowed us to define various functional spaces (Fig. 36).

62The main streets, especially the so-called decumanus and cardo, follow the terrain and, above all, connect the natural approaches to the plateau; they do not seem to have changed over time (in green in Fig. 29). Although a certain regularity can be observed in the streets, which is mainly reminiscent of a herringbone pattern, the streets nevertheless do not have a strict orthogonality. This peculiarity seems to be rather typical of Etruscan and Italic settlements and can be compared with the situation in Veii[124] and Tarquinia[125]. In Gabii, the central axis follows the morphology of the volcanic crater of Lago Castiglione and creates a quasi-orthogonal raster with the secondary streets, which was developed ex-novo at the end of the 5th century B.C. and then renewed several times[126]. In Tarquinia, preliminary excavation results indicate that the street system was laid out in the late 6th or early 5th century B.C.[127]. A fundamental redesign with the creation of an orthogonal street network can be traced in Veii in the area of Piazza d’Armi around the middle of the 7th century B.C.[128]. These examples suggest a reorganization that has radically changed the perception and division of the physical space of the cities and their street system. It is not yet possible to date the red system in Vulci and although it appears to respect both temples, further excavations will clarify whether it could possibly belong to a wider phenomenon of monumentalization of urban spaces in central Italy in the late Archaic period. One could, however, relate these interventions to demographic or social changes[129] and it will be one of the goals of this project to define the challenges or transformations that led to a particular urban solution and to the reshaping of the street grid.

63The sacral topography of Vulci is strongly connected to traffic routes, both in and outside of the city, and to the gates[130]. Important sacred spaces intra muros are, to name just a few: the votive deposit at the porta nord[131], the sacello di Ercole[132] and the new sacred district, which can all be found along the main urban streets. Both monumental, late archaic temples are oriented towards the so-called decumanus and occupy a central elevated position, facing south and ideally towards the sea (even if it was not visible from the temples). Their monumental silhouette may have dominated the cityscape of Vulci and have significantly shaped the visual and ritual perception of the city – also from a diachronic perspective. Current evidence suggests that both temples were erected at roughly the same time in the Late Archaic period[133]. In Roman times, the smaller cult building in the south was added to the ensemble, an Augustean renewal has been assumed for the tempio grande[134] and at some point, in the Early Imperial Era, at the latest, the »new« temple fell into ruin.

64The discovery of a new temple next to the tempio grande hints at a complex ritual landscape (Fig. 35). Adding a new urban temple to the core of the cityscape of Vulci is reminiscent of other sacred districts with two adjacent temples in Etruscan cities, such as those in Volterra [135] and Pyrgi [136] or the sanctuary of S. Antonio in Cerveteri [137] or in Marzabotto [138]. These contexts show a diversified architectural spectrum which, nevertheless, share certain characteristics. The temples have, for example, the same orientation towards the sea or a main street (Fig. 35), thus underlining the relationship between sacral areas and traffic routes and in so doing, structuring both urban and rural spaces. Both temples in Caere date to the late 6th/early 5th century B.C.[139], the temple of Tinia and Uni in Marzabotto were built just a few decades apart, around the end of the 6th/beginning of the 5th century B.C.[140], while the temples of Pyrgi and Volterra were planned in a different time period[141], yet still respect the plan and the orientation of the former constructions. As in the other examples, it seems plausible that the two monumental temples of Vulci were coeval and dedicated to different gods. This ensured the religious representation of a larger part of the population, thus preserving social differences and preferences, creating a sense of inclusion for the wider public and contributing to social cohesion.

65Through a multidisciplinary approach and the use of state-of-the-art geophysical prospection techniques, it was possible to propose a new reconstruction of the cityscape of Vulci, covering the entire urban area north of the so-called decumanus and providing a glimpse into the city's complex palimpsest, which spans more than 1,500 years.

66The use of two complementary methods, magnetometry (Fig. 37) and GPR (Fig. 38) in Vulci can be confirmed as thoroughly successful, providing evidence of several structures and functional areas and, in some cases, proposing a provisional chronological sequence of roads and structures and evaluating the way in which they interact and relate to one another. It was possible to identify part of the defensive system, as well as residential and productive areas. Most interesting, however, is the new sacred district that emerges near the tempio grande, where another monumental temple and a smaller building occupy such a prominent place in the centre of the city. Thus, we had to change our picture of the spatial organization of spaces, physically divided and organized via street systems, and were also able to highlight how functional clusters can be integrated and interwoven into an urban system. The combination of the two geophysical methods, together with archaeological excavations, is especially effective, revealing complementary features. While the use of large-scale magnetometry made it possible to trace the urban structure of the northern area of Vulci, the GPR made additional structures visible and showed a high level of detail. Furthermore, the discovery of hitherto unknown monumental buildings and the detection of various functional areas indicate the potential of breaking away from the small-scale, site-oriented approach, and the value of using geophysical survey technologies to identify larger areas or even whole cities.

67The excavation of the temple confirmed the results of geophysics, laid the foundation for a further systematic refinement of geophysical technologies, and connected the spatial layout with its diachronic development. The integrated use of non- or minimally invasive methods, as well as targeted excavations, which are also planned for the future, promise to decisively enrich our picture of the ever-changing cityscape of Vulci. The aim must be to obtain a picture of a versatile city in which different functional areas are intertwined, are related to larger archaeological and historical developments, and evolve over the centuries as breathing organs of the city structure.

68Our first and heartfelt thanks go to Simona Carosi and Carlo Casi and to the institutions they represent – the Soprintendenza Archeologia, Belle Arti e Paesaggio per la provincia di Viterbo e per l’Etruria meridionale and Fondazione Vulci, Parco Archeologico e Naturalistico di Vulci, respectively – for believing in the project from the very beginning, and for their tireless and constant support. Gratitude is owed, moreover, to the Fritz Thyssen Stiftung which generously funded the project over the years (Az. 20.19.0.028AA and Az. 20.21.0.0055AA) making both the geophysical prospections and the first excavation campaign possible; similarly the Gerda Henkel Stiftung generously supported our second excavation campaign 2022 (AZ 20/V/22). We wish to thank the departments of Archaeology of the Universities of Mainz and Freiburg for their help and support; in particular, we would like to thank the Top-level Research Area 40,000 Years of Human Challenges: Perception, Conceptualization and Coping in Premodern Societies of the JGU Mainz and the members of the Thematic Area 3 – Urban Agglomerations for the support, the stimulating inputs and fruitful discussions. For the helpful discussion regarding the defensive system, we warmly thank Sophie Helas. A warm and fond thank-you goes to the staff of the archaeological park for their invaluable support. Heartfelt thanks also go to the other members of the Eastern Atlas team, who undertook the survey in September 2020: Annika Fediuk, Ronald Freibothe and Henning Zoellner. Last but not least we wish to warmly thank our team, who have enriched the first two of many future excavation seasons: Melina Angermeier, Manuela Battaglia, Julian Brosius, Julius Bussilliat, Quirin Eimer, Cai Ellrich, Roland Frank, Jessica Hahn, Eliza Hernández Osorno, Felicitas Kandler, Maki Müller-Quade, Katharina Oppelland, Laura Rausch, Thomas Schmidt, Laura Schmitt, Philipp Schug, Nadja Schulz and Alexandra Wiesbeck-Klein.

For a long time, Etruscan cities, their physical appearance, diachronic evolution and cultural and social figurations have largely been neglected in archaeological research. Only in recent years has this begun to change thanks to the systematic application of new methodological approaches. By using a combination of non-invasive, geophysical prospections and targeted excavations at neuralgic points, the Vulci Cityscape project aims to examine the cityscape of Vulci and its transformation over the longue durée. Geophysical surveys conducted in 2020 on 22.5 ha north of the so-called decumanus resulted in a new and more complete plan of this part of the city, identifying different functional areas, differentiating street systems and revealing the complex historical palimpsest of the urban structure. Among the functional areas, a new sacred district to the west of the tempio grande, including a new monumental, late Archaic temple stands out. Not only do the results improve our knowledge of the urban layout of Vulci, but they also shed new light on Etruscan urbanism in general.