Sensing the Cultural Heritage from Above. The Case from Cyprus

Abstract

This chapter addresses the different remote sensing methodologies that have been applied for the study of the Cultural Heritage in Cyprus. Ground based geophysical prospection, aerial and satellite remote sensing, in tandem with soil analyses of cores, have been applied for the mapping of the archaeological sites and the reconstruction of the archaeoenvironment, but also for addressing issues related to the risk assessment of sites and monuments. Taking into account the different geological conditions of the island and some of its peculiarities (such as metamorphic and iron-rich geological formations), the success of these methods varies significantly. The past experiences can be used as a guideline for the wider and more successful application of the remote sensing techniques.

1 Introduction

Cyprus is one of the first localities in the Mediterranean world which made use of non-invasive research for archaeological studies. Since the 1970s, several studies employed geophysical surveys, remote sensing, and GIS application to investigate archaeological settlements, necropolis, and landscape. Other studies aimed at restituting the environmental context and its evolution in relation to ancient human occupation.

The range of geophysical methods developed to investigate the archaeological sites included ground penetrating radar (GPR), magnetometry, earth resistance mapping and electrical resistivity tomography (ERT), electromagnetic induction (EMI) and chemical analysis. Some specific limitations have been identified for each method and some archaeological features were better delineated when several methods were employed together and when the soil conditions and geological background allowed. Concerning the reconstruction of environmental evolution, cores, natural outcrops, and trenches were studied using multidisciplinary methods that included granulometry and magnetic susceptibility measurements associated with dating.

On a larger scale, aerial imagery and satellite remote sensing employing a large variety of space sensors have been used for the detection of archaeological sites and cultural heritage management, sometimes in tandem with ground-based prospection techniques. In many cases, the above have also been incorporated in Geographical Information Systems (GIS), and spatial analysis was used for a further assessment of the settlement patterns in different periods of antiquity.

The aim of this chapter is to present a condensed review of what has been done in terms of geophysical surveys, coring, and satellite remote sensing for archaeological applications in Cyprus until today (Fig. 1).

Fig. 1

Archaeological sites and case studies mentioned in the text.

2 Environmental Background

Cyprus is the third largest island in the Mediterranean. The island belongs to the Mediterranean climate zone and therefore, experiences mild winters and hot dry summers. The wet season extends from November to March, with most (approx. 60%) of the rain falling between December and February (Pashiardis, 2002). The Troodos Mountain rage represents the dominant source of hydrologic activity where major rivers of the island originated. Among these we can cite the Pediaios and the Gialias that originate from the eastern part of the Troodos Mountain, drains the Mesaoria plain and end within the Mediterranean Sea in the eastern part of the island. Approximately two third of the island is covered with mountains. Its topography, which is related to the geological history of the island is mainly composed of two mountain range (Zomeni, 2012). The Pentadaktylos (Keryneia) mountain range is located to the north and is composed of recrystallised sedimentary deposits (limestone, dolomites and marbles). The southern part of the island is characterised by the Troodos Mountain range. It is composed of volcanic and metamorphic rocks related to the ophiolithic sequence. In the southwestern part of the island, the Mamonia formation, has been formed during Upper Triassic-Cretaceous and is composed of sedimentary rocks and basalts. Circum Troodos geological formations are mainly composed of marine and continental deposits. Between the two main mountain range, the Mesaoria plain is composed of conglomerate formation related to Pliocene and Pleistocene alluvio-colluvial fan (Harrison et al., 2013).

In a few cases, the metamorphic geology and the iron-rich deposits have created problems with respect to the efficient application of remote sensing techniques, especially in magnetic prospection. Furthermore, the compacted clay soils that exist especially in the valleys has created a strong compacted soil context that does not allow the clear identification of subsurface targets through the GPR surveys. The above are further attenuated from the deep ploughing and the intensive agricultural practices that sometimes have scratched the cultural horizon, diffusing in a large degree the past anthropogenic traces, which creates certain difficulties in the identification of the underneath architectural relics by the airborne and satellite sensors.

3 Ground Based Geophysical Surveys and Geochemical Analysis

The existence of a physical heterogeneity between the underlying archaeological targets and the surrounding natural sub surface (soil, geological bedrock, etc.) is the basic principle of main types of geophysical survey (Aitken, 1974; Nishimura, 2001; Scollar et al., 1990). Mineral and chemical composition of the sub-surface can be obtained respectively from measurements of the magnetic susceptibility and pXRF, in situ or from soil samples analyses in the lab. When cross combined with excavation, the shape, the geometry and the depth of these anomalies can be related to specific archaeological features. However, outcrops of the geological bedrock, intense sources of noise or small contrasts between the archaeological features and their soil environment can hide or make invisible the targets.

For more than 50 years, some of the most important archaeological sites of Cyprus have been studied through ground-based prospection techniques (Aitken, 1971; Fischer, 1980; Hesse & Renimel, 1978; Parhas et al., 1979). Limitation of geophysical surveys are mostly linked to climate (high temperatures) and geology. In the eastern part of the island, GPR measurements were strongly influenced by clay and salt rich sediment from Larnaca Salt Lake. Arid conditions in the western area of Cyprus induced very low readings in the resistivity surveys (Graham et al., 2013; Boutin et al., 2013; Gibson et al., 2013). Copper slag piles formed from copper extraction processes during Chalcolithic to Bronze Age period (and continued until recent historical period) together with very magnetic geological formations prevent the extensive use of magnetic methods near the Troodos Mountain range. Although some limitations still exist, several techniques have been implemented in order to depict the archaeological settlements and necropolis of the island from the Neolithic to the Medieval period. The most often used method remains the GPR, magnetic and electrical surveys. The ERT technique and EMI method are still only applied at a limited number of sites in Cyprus. Most of the investigations involved more than one method of prospection in a “manifold” geophysical approach (Sarris, 2013; Kalayci et al., 2017). Although the geomorphological context can be considered as a limitation, some geophysical survey techniques allowed to locate the sectors of extensive distribution of architectural relics with a better resolution of the archaeological filling. Best examples are illustrated at Klimonas—Agios Tychonas (Benech et al., 2017a) and at Kition (Benech et al., 2017b).

Just to present a few representative examples, excavations at the site of Klimonas—Agios Tychonas (Fig. 1) yielded the remains of a Pre-Ceramic Neolithic A (PPNA) village with mud-brick circular structures dated to the beginning of the ninth millennium BC (Vigne et al., 2017). Geophysical surveys (Fig. 2a) allowed to highlight structures that are more complex to identify due to the low contrast between the architectural elements and the surrounding soil. Although the magnetic response was weak in amplitude and small in size, several discrete structures like pits and raw earth architectural remains have been identified. The EMI survey was rather indicative of the geological substrate due mostly to the rough geomorphology and the dry environment (weak signal registration). The recognition of the geophysical signatures met a number of difficulties due to the rough geomorphology, the outcrops of the bedrock, the lack of sufficient thickness of soil, the dry environment (weak signal registration) and the intensive soil corrosion phenomena (lack of signal distinction) (Benech et al., 2017a).

Fig. 2

Example of archaeo-geophysical surveys. (a). Klimonas—Ayios Tychonas. Left: Magnetic susceptibility map (in 10⁻⁵ USI) resulting from an electromagnetic survey with a CMD Mini-Explorer for a 71 cm coil gauge (drawing A. Tabbagh; topographic base R. Touquet). Right: Magnetic susceptibility map resulting from an electromagnetic survey with a CMD Mini-Explorer for a 118 cm coil gauge (drawing A. Tabbagh; topographic base R. Touquet). The geophysical anomalies are mostly linked to the geological bedrock rather than to architectural remains or ancient human activities. However, high values (in red to orange on the map) are obtained on the terraces of the archaeological site of Klimonas. (From Benech et al., 2017a, b). (b). Dromolaxia—Vyzakia (Hala Sultan Tekke). Magnetometer map of the archaeological site. The colour scale is presented in Black (negative anomaly) and white (positive anomaly). The red rectangle indicates the area of excavation CQ4. The interpretation of the magnetic survey was confirmed by excavation and highlight several buildings organised in large compound. (From Fischer et al., 2020). (c). Kalavasos—Ayios Dhimitrios. Time slice at depth of 60–80 cm. Light colours indicate high amplitudes and dark colours low amplitudes. Black rectangular voids are locations of extant olive trees. B. = building number (drawing by T. Urban and K. Fisher). The interpretation of the GPR survey was confirmed by excavation and highlight several buildings organised in large compound. (From Fisher et al., 2019).

Dromolaxia—Vyzakia (Hala Sultan Tekke) is a large Bronze Age city located on a Quaternary alluvial fan near the palaeo-depression of the Larnaca Salt Lake (south-east coast of Cyprus, Fig. 1) (Devillers et al., 2015). Since the first settlements ca. 1600 BC, the site was probably the most important harbour in the entire Eastern Mediterranean during Late Bronze Age (ca. thirteenth–twelfth c. BC). Around 1200–1150 BC, several layers of destruction have been also identified with the “crisis years” at the end of the Bronze Age in the Mediterranean. The extensive surveys using magnetometer (Fig. 2b) indicated large architectural compounds intersected by streets and tombs from the necropolises (Fischer et al., 2020; Trinks, 2015). The mineralogical and chemical analysis on soil samples allowed to highlight distinct characteristics between interior and exterior of structures, which can be related to by-products of domestic or industrial activities (Cuenca-Garcia et al., 2015; Hafez et al., 2017). GPR survey was limited to the uppermost parts of the archaeological and natural deposits due to the salty clay-rich soil inducing a strong electromagnetic radar attenuation.

Kalavasos—Ayios Dhimitrios (Fig. 1) is a significant regional centre, well-positioned site at a confluence of routes that link the eastern part of the island to the west, the copper mines from the Troodos Mountains to the north and the coast to the south (South, 1980). Previous archaeological studies suggest that during the Late Bronze Age period (ca. 1450–1200 BC), the site likely covered 11 ha, built around an urban centre (Keswani, 1993; South, 1997; Todd, 2004). The site’s administrative centre is characterised by a 30.5 × 37 m court-centred monumental structure with ashlar masonry dedicated to the production and storage of olive oil and elite feasting activities (South, 1997; Fisher, 2009). The GPR survey (Fig. 2c) detected a complex of large quadrangular structures with apparent partitioned spaces in the South and in the West part of the site (Fisher et al., 2019; Rogers et al., 2012; Urban et al., 2014). Complementary EMI survey identified an area of remnant salts from organic matter in a central room or court that could possibly be related to by-products of feasting activities (Fisher et al., 2019). The same building exhibits an area of high magnetic susceptibility that might be related to some intense burning processes such as cooking. The lack of contrast between the magnetic properties of the calcareous soils and the limestone and sandstone building materials that were used to construct Late Bronze Age buildings limited drastically the magnetometry survey (Fisher et al., 2019).

Kouklia—Palaepaphos (south-west coast of Cyprus) is recognised as an UNESCO (United Nations Educational, Scientific and Cultural Organization) World Heritage site (Fig. 1). It has been one of the most significant sites of the island with a constant human presence from the Chalcolithic period to recent times (Maier & von Wartburg 1985). During Late Bronze Age (around the thirteenth c. BC), Kouklia Palaepaphos grew into one of the island’s first regional polities and during the Iron Age, the site is considered as one of the ten blooming city-kingdoms of Cyprus (Iacovou, 1994, 1999; Maier, 1999). Along the fourth c. BC, when port facilities and administrative functions were transferred to Nea Paphos (18 km to the West), the urban landscape began to shrink. Only the open-air sanctuary to Aphrodite and its direct environs continued to receive attention during Hellenistic and Roman eras until the advent of Christianity. Previous hypotheses of the extension of the urban sectors and the fortification walls at Kouklia—Palaepaphos were confirmed and even rejected by systematic and extensive geophysical surveys in tandem with GIS spatial analyses (Iacovou et al., 2009; Sarris et al., 2005; Sarris & Papadopoulos, 2010; Stamatis et al., 2007). The multi-method survey (GPR, magnetic and electric) was able to identify monumental, domestic buildings and allowed to reconstruct section of the temenos wall of the sanctuary area. Some anomalies that were suggested by all methods proved to have been caused by geological processes (iron enriched bedrock or residues of past lightings, Fig. 3a).

Fig. 3

Supplementary examples of archaeo-geophysical surveys. (a). Laona region of Palaepaphos (Kouklia). Results of the magnetometer survey. Measurements with both Bartington G601 and Geoscan Research FM256 came show several extreme magnetic anomalies (>+/−3000 nT, Black (negative anomaly) and white (positive anomaly)) of relatively large dimensions (~5–10 m). Subsequent excavations did not result any kind of recent or ancient anthropogenic feature. Soil susceptibility analysis indicated a relative high level above the specific anomalies. The intensity of the magnetic anomalies and their amorphous shape is most probably related to the effect of lightings that hit the area. This may be related to the rich copper-bearing outcrops of bedrock that exist in a low depth below the ground surface. (From Sarris et al., 2014). (b). Nea Paphos. ERT results of the underwater survey at Kato Paphos. 3D isosurface of the resistivity values more than 5.8 Ohm-m showing the extend of the submerged wall structure related to the antic harbor up to 1.5 m below the seabed. (From Simyrdanis et al., 2017). (c). Nea Paphos—Neapolis. Result of the magnetic and GPR survey. The colour scale is presented in Black (negative anomaly) and white (positive anomaly) for the magnetometer survey and the intensity of the GPR signal is highlighted by hot (yellow to red) colours. Potential candidates to represent architectural features registered on the magnetic and GPR measurements. (From Sarris & Papadopoulos, 2019).

Nicosia (Fig. 1), the capital of Cyprus since the tenth century (Papacostas, 2012), is located in the Mesaoria Plain, at the central part of the island. Ancient Nicosia has been occupied since the Late Chalcolithic period (Pilides, 2004; Hermon et al., 2014) but the modern city completely overlies architectural remains assignable to the early Christian period until the sixteenth c. AD. The Venetian fortification and the moat (1489–1570) have a circular shape containing eleven pentagonal bastions and three gates (Bakirtzis, 2017; Grivaud, 1992; Jeffery, 1907; Panciera, 2010; Violaris, 2012). The multi-method geophysical survey (ERT, EMI and GPR) allowed to reconstruct a section of the Venetian wall and of D’Avilla bastion (Cozzolino et al., 2020). Best results were obtained with the ERT method as EMI and GRP measurements were mostly related to modern underground structures (pipe, electric cabin). Some resistive anomalies detected by the ERT were verified by excavation and attested to a perfect correspondence between the geophysical previsions and the Venetian wall found in the subsoil (Cozzolino et al., 2020).

The archaeological site of Nea Paphos (southeastern coast of the island, Fig. 1) revealed important edifices related to the Classic Antiquity (House of Dionysos, the House of Orpheus, the Villa of Theseus and the House of Aion; the Agora, composed of an Odeon and the Asklepieion; the Theatre and the necropolis of the “Tombs of the Kings”) and is inscribed on the World Heritage List of UNESCO since 1981. The city was founded at the end of the fourth c. BC when port facilities were transferred from Kouklia—Palaepaphos (Bérard, 1954; Michaelides, 1991; Papageorghiou, 1983). While disastrous earthquake ravaged most important cities on the island, the city became the central administrative centre of the Ptolemaic kingdom on the island around the third c. BC. The city began to shrink after the Arabs incursion of the seventh c. AD but continued to play an important role in the island during following centuries, especially during the Byzantine (eleventh c. AD) and Medieval period (around 1500 AD) (Altinok et al., 2011; Fokaefs & Papadopoulos, 2007). The ERT method implemented at the Hellenistic to Roman harbour of “Kato Paphos” (Fig. 3b) was able to provide solid evidence on possible building walls buried 2 m under the seafloor (Simyrdanis et al., 2017; Papadopoulos, 2021). In the Agora, preliminary results of the magnetic survey realised in 2015 indicated several rectangular anomalies that have been related to diamagnetic slabs of limestones, probably used for the pavement of the pathway. High amplitude magnetic signals were related to some granite column (in the form of concentrated dipoles) and low amplitude readings to architectural remains, filled pits, double canal made of limestone and clay pipes (Seifert et al., 2020). Around the city centre, two survey aimed at identifying ancient architectural remains. Multi-technique survey (GPR and magnetometry, Fig. 3c) demonstrated that most of the area is without significant ancient occupation (Sarris & Papadopoulos, 2019). Some exceptions are consisting of a number of linear anomalies and round features. The particular features are obvious in both the magnetic and the GPR data, although the magnetic measurements are obscured from high levels of noise (Sarris & Papadopoulos, 2019). A second study investigated four different sectors in the archaeological park and in the modern city (Benech, 2014). All sectors were polluted by superficial disturbances from neighbouring modern construction. The results delivered interesting magnetical anomalies whose organisation and orientation are in good correlation with the Roman villa discovered during excavation in the area by the French mission.

4 Coring and Reconstruction of Archaeoenvironment

The geomorphology and the landscape of Cyprus is mostly linked to the geological bedrock and to the climate of the island. Human activity is considered as a factor of transformation of the landscape (canals, rock extraction, urbanisation, etc.) in dynamic equilibrium with climate changes. The proximity of the main archaeological site with coastal, lacustrine and fluvial sectors is in relation with their evident benefit for the ancient societies’ activities. As a counterpart, climatic events seemingly influenced on socioeconomic changes by impacting sowing and growing seasons and irrigation capacity.

Landscape reconstruction of these areas is obtained through the analysis of stratigraphic profiles from natural context or from core extraction. Several parameters are used on regularly distribute samples extracted along the profiles in order to reconstruct the ancient landscape and human activity (Fig. 4). Magnetic susceptibility measurements (low and high frequencies) can distinguish different sediment sources (Ghilardi et al., 2015; Sarris, 1992; Vella et al., 2019), soil formation (Dearing et al., 1996), firing events (Oldfield & Crowther, 2007) and pottery production or other workshop activities (Dearing et al., 1996; Jordanova et al., 2003). Grain size determinations (laser analysis for the fraction below 2 mm and sieving for coarser fraction), organic matter and CaCO₃ content (loss on Ignition) are generally used to quantify the depositional and erosional processes (Vella et al., 2019). Pollen and charcoal analysis contributes further to the reconstruction of the composition of the vegetation. Finally, the chronological control of the environmental evolution is obtained mostly through AMS radiocarbon (charcoals bones and shells) and OSL (sherds) dates.

Fig. 4

Paleoenvironmental reconstruction based on cores extraction and pollen and sedimentological analysis. Dates are obtained through ¹⁴C and OSL dating

Although well developed at the scale of the Mediterranean, multi method analysis of subsurface samples extracted from coring is still limited in Cyprus. The most studied area is represented at the south-eastern sector of the island in the vicinity of the Larnaca Salt Lake (Fig. 4). Two major archaeological sites are excavated for more than 100 years by French, British, Swedish and Cypriot groups. The particular studies at Dromolaxia—Vyzakia (Hala Sultan Tekke) (Bronze Age, ca. 1600 BC to ca. thirteenth–twelfth c. BC) and Kition (Cypriot Iron Age to the Hellenistic period) indicated that the area was continuously inhabited for the last 3500 years. The Larnaka Salt Lake delivered two continuous undisturbed sediment cores, with chronological (¹⁴C dates), sedimentological, and paleoecological (pollen analysis) correlations allowing to create a unique sequence covering the period ~4000 ± 20 BC to 1500 ± 50 cal year AD (Kaniewski et al., 2013). Several cold and dry periods were recorded at 3.2 ka BC, 2.2 ka BC, 1.2 ka BP, 800 AD and 1200 AD in concordance with cooler phases in Europe (Kaniewski et al., 2020). Some other cores in the sector could identify several natural channels between the Salt Lakes lagoon and the open sea. Coastal progradation, associated to siltation of important naval routes of communication, participated to the abandonment of the large sheltered anchorage of Dromolaxia—Vyzakia (Hala Sultan Tekke) during the early twelfth c. BC. The canal, which was excavated during the later second millennium BC, could be a response to the siltation during the Iron Age. At Kition, cores extraction and paleoenvironmental reconstruction are systematically employed for the last 40 years. At around ca. 900 cal. BC and 500 cal. BC (Iron Age) the site overlooks a sea bay evolving into a lagoon and then into an increasingly enclosed marsh land following the formation of a coastal bank (Bony et al., 2016; Gifford, 1978; Morhange et al., 2000; Nicolaou, 1976). Thus, the military port was founded in the classical period in a lagoon environment, the coastal bank of pebbles having favoured the installation of a port activity in the most protected part of the lagoon. To the west of the Larnaka Salt Lake, within the Tremithos Valley, 8–10 m of fluvial sediments are recorded between ca. 4800 cal. BC (Sotira Neolithic Culture period) and ca. 2500 cal. BC (Late Chalcolithic) (Ghilardi et al., 2015). Around 3200 cal. BC, the river incised into the alluvial formation, probably in relation with flash-flood events during cold and dry climatic events centred on 3.2 ka BC (Bar-Matthews et al., 1997; Roberts et al., 2011).

The Gialias valley is the second area investigated to study the Holocene fluvial reconstruction in Cyprus (Devillers et al., 2006). The geomorphological study of the median sector of the valley has identified a first alluvial terrace which presents several paleosols dated to 9300, 7400, 4400, 3800 cal. BC. Downstream, marine vases fill the lower Famagusta valley. All the watershed region experienced very high sedimentation rate with a negligible human influence on detritism. Between 3500 and 2000 cal. BC, the first major incision phase was linked to low water intake and/or relatively high temperatures. This event can be reasonably related to the cold and dry periods of the 3.2 ka and 2.2 ka BC events identified by Kaniewski et al. (2020) at Dromolaxia—Vyzakia (Hala Sultan Tekke). Downstream, the presence of shell sands testifies to the establishment of a coastal barrier. Between 2000 cal BC and 1250 cal. AD, a new significant alluvial period and two phases of pedogenesis are identified (between 4400 cal. BP and 1400 cal. BC; 800 cal. BC). During the Middle Ages (300–1250 cal. AD), a new fluvial incision affects the middle valley corresponding to the cold and dry event of 800 AD identified by Kaniewski et al., 2020 at Dromolaxia—Vyzakia (Hala Sultan Tekke). The sedimentation rates of the Frankish, Venetian and Ottoman periods (1250–1900 cal. AD) are important and characterised by sands with lenses of pebbles. The raising of the alluvial floor between (4 and 5 m) causes significant changes in the morphology of the alluvial plain during the LIA. The developments related to the fluvial network (mills, bridges, dams, etc.) then become unsuitable and are quickly become buried. Incision 3, could not be precisely dated but could be attributed to the first half of the twentieth century.

In western Cyprus, several studies aimed at the reconstructing the chronology of the alluvial erosion and filling phases since Holocene. The Khrysokhou, Ezousas, Dhiarizos, and Xeropotamos River terraces have been analysed through grain size and soil analysis (magnetic susceptibility, OM content and pH), the chronological control of samples been obtained with OSL datings (Deckers, 2005). Although early to mid-Holocene fluvial sediments may have been eroded in most cases, Byzantine to modern period have been more documented. Within the Khrysokhou valley (PITSI/Section TA-TB), two period erosion phases were identified around 574 AD and 920–1571 AD. The Ezousas valley presented evidences of possible early to mid-Holocene alluviation (EZA) and flood events around 1050–1060 AD (EZA and EZD respectively), 1297 and 1329 AD (EZG). Near Kouklia—Palaepaphos archaeological site, the lower valley of the Dhiarizos River delivered 3.1 m of rounded gravels overlied by fine silt (KOL1). OSL dating is estimating the age of that fluvial deposits at 1600 AD. Farther inland, fluvial sediments have been dated to 1200 AD (MB) and 1400 AD (PR1). Upstream, OSL datings on sherds attest to the presence of a sediment deposition shortly after 920 AD (KIS1). At the mouth of the Xeropotamos River, the sequence of deposits is dated sometime after 1760 AD (XA-XE). Finally, the Mitsero Basin in western central Cyprus showed a very similar absence of fluvial sediments dating between the early Holocene and the Medieval period (Given & Knapp, 2003). The incision phase identified in the Khrysokhou valley can reasonably be related to the other one identified between 300 and 1250 cal. AD at the Gialias valley (Devillers et al., 2006) and the cold/dry event at 800 AD in Larnaka Salt Lake (Kaniewski et al., 2020). Medieval to modern fluvial deposits (1297–1760 AD) are most probably linked to the LIA event.

5 Satellite Remote Sensing, Aerial Photography, and Ground Spectroscopy

Scientific literature related to satellite processing for supporting archaeological research (Luo et al., 2019) has been growing around the globe. This increase implementation is directly linked with the availability of meter and sub-meter satellite sensors, which took place at the end of the twentieth century, as a result of the release of the first commercial high-resolution IKONOS satellite sensor in 1999. Since then, several other high-resolution commercial satellite sensors have been set into orbit (Agapiou & Lysandrou, 2015).

In Cyprus, satellite remote sensing, aerial photography and ground spectroscopy have been only recently introduced. Hereunder, are briefly presented some examples of archaeological projects that implemented in their research process satellite and/or aerial investigation methods, tools, and data. Also, some examples of basic scientific research fulfilled within the framework of funded research projects are given. All presented examples were accomplished as part of the multidisciplinary research group of the Archaeology and Cultural Heritage section (ARCH), Remote Sensing and Geo-environment research Lab, established at the Cyprus University of Technology.

One of the first integrations of multitemporal satellite archives with higher resolution aerial datasets was implemented under the Palaepaphos Digital Atlas (2002–2003) and the Palaepaphos Urban Landscape Project (PULP) (2006–today) (Iacovou et al., 2009; Iacovou, 2008). For these projects, high-resolution multispectral satellite data like the GeoEye, IKONOS and QuickBird, along with multi-temporal orthophotos provided by the Department of Land and Surveyors of Cyprus, were processed. Also, compressed RGB images from the Google Earth digital Globe were extracted and elaborated. At the same time, the above-mentioned geo-dataset was used to provide background information on the landscape through the GIS environment.

An integration of archive aerial images with recently acquired high-resolution satellite datasets was carried out in the case of Graz Amargeti Survey Project, directed by Dr. Gabriele Koiner and Dr. Gabriele Ambros (For preliminary reports refer to (Amargeti Survey Project, 2021) and (Graz Amargeti Survey Project, 2021). Specifically, a WorlView-2 multispectral image and archive aerial orthophotos provided by the Department of Land and Surveyors of Cyprus (1963, 1993 and 2014) were used. Linear features (cropmarks) were extracted in specific plots of the case study area before the ground geophysical investigations. Despite the limited spatial resolution of the aerial data, results were found very encouraging. These, along with the geophysical prospection’s outputs, will be used to support and guide future archaeological field investigations.

A recent project concerned the mapping of the ancient monuments (declared as such and protected by the Antiquities Law), in the Paralimni Municipality (Agapiou et al., 2020). To geolocate the monuments and sites under question, archive aerial images and cadastral maps (provided by the Department of Land and Surveyors of Cyprus), were used. Given the recent land-use changes in the area, the exact geolocation of the protected monuments was achieved through the interpretation of aerial photos, cadastral maps, and archaeological records. As the cadastral plans of Cyprus have changed over the last decades, and the plan, sheet and plot numbers have been modified, the detection of the monuments—especially of those that have been declared protected decades ago—through the archival, archaeological information required confirmation from the historic aerial datasets to match the plots with the protected zones of the monuments.

Another recent example was the investigation of the Xeros valley, under the auspices of the Settled and Sacred Landscapes of Cyprus (SeSaLaC) project (directed by Associate Prof. Thanasis Vionis), through aerial photographs. Despite the limitation of the spectrum resolution (analysis was conducted only in the visible part of the spectrum), the results were found very promising as they were aligned with the results of the geophysical prospection surveys. Several hot spots were identified and mapped in GIS environment through image enhancement techniques, including histogram stretching and filtering. Based on the findings, hot-spot analysis and clustering was applied to group the detected archaeological proxies.

A more methodical approach towards the integration of satellite, aerial remote sensing and ground spectroscopy for archaeological studies, was initiated in 2012 with the first related PhD thesis (Agapiou, 2012) the core of which was the use of satellite, middle range and ground remote sensing techniques, along with geoinformatics towards archaeology and built heritage monuments. In Agapiou (2012), the use of satellite remote sensing datasets was investigated to detect cropmarks, which were used as archaeological proxies for subsurface archaeological remains. Such proxies are detected due to their unique spectral signatures and the distinct contrast that they provide in relation to the existing cultivations. The optimum temporal and spectral resolution for supporting these types of investigation in the Eastern Mediterranean region was identified (Agapiou et al., 2013). The processing of hundreds of ground spectral signatures obtained from simulated test fields indicated that the 760 and 900 nm spectrum regions are the best wavelengths to support the interpretation of cropmarks and their semi-automatic extraction from the images. It was also found that the period between the mid-March and the beginning of April is the optimal temporal window for observing and detecting cropmarks over cultivated areas in Cyprus.

A challenging task was the study of the optimum spatial resolution for supporting landscape archaeology, especially in areas with spectral heterogeneity. The optimal spatial resolution (OSR that is the ground spatial resolution or the pixel size of the images), for two different cases studies, a simulated archaeological environment in Alampra village test field and the archaeological site of “Nea Paphos”, were investigated (Agapiou, 2020). The local spectral variance of a given area of interest (e.g., archaeological proxy) is minimised without losing key details necessary for an adequate interpretation of the cropmarks. The spectral range was limited to the visible and near-infrared part of the spectrum (400–900 nm). The OSR was estimated for each spectral (RGB), and near-infrared bands. The study was also expanded to include vegetation indices, such as the Simple Ratio (SR), the Atmospheric Resistance Vegetation Index (ARVI), and the Normalised Difference Vegetation Index (NDVI). Based on these findings, the OSR for the above case studies was defined (Fig. 5). The outcomes indicated that the OSR could minimise the local spectral variance, thus minimising the spectral noise, and, consequently, better support image processing to extract archaeological proxies in areas with high spectral heterogeneity.

Fig. 5

(a) Simulated simple ratio (SR) datasets pixel size 1; (b) simulated SR datasets pixel size 2; (c) simulated SR datasets pixel size 3; (d) simulated SR datasets pixel size 4; (e) simulated SR datasets pixel size 5; and (f) simulated SR datasets pixel size 10. (Agapiou, 2020)

In parallel, another study was performed in environments with spectral heterogeneity. In these areas, interpretation and detection of cropmarks can be problematic even after applying sophisticated image enhancement analysis techniques due to the phenomenon of mixed pixels. To overcome this problem an image-based methodology over specific case studies in Cyprus where the vegetation is suppressed following the “forced invariance” method, was proposed (Agapiou, 2019). The promising results of this study were evaluated in the archaeological site of “Nea Paphos” in Cyprus using a WorldView-2 multispectral image (Fig. 6).

Fig. 6

Vegetation suppression (NIR-R-G composite, top) and pansharpened multispectral image (NIR-R-G composite, bottom). (Agapiou, 2019)

The use of aerial photography as an investigation tool for archaeological purposes in a more systematic basis has only lately started in Cyprus. A recent study gathered all known archive and new aerial photographs over Paphos area, to examine its burial grounds (Lysandrou & Agapiou, 2020). In that study, specific aerial datasets were elaborated, and blended with archaeological and topographic data, investigating the Hellenistic (-Roman) eastern necropolis of Nea Paphos boundaries and tombs. Some of the aerial data were produced before the excavation of the site, and therefore were of valuable utility. This investigation was based on metrics extracted from known Hellenistic tombs of the necropolis (Lysandrou, 2020). Thereafter, more archaeological features, possible other tombs, were detected and interpreted through the aerial images (Fig. 7). The use of multitemporal archives enabled the reconstruction of the landscape of Paphos before the modern urban expansion, revealing at the same time various soil and cropmarks that share common characteristics. All archaeological remains and proxies have been introduced into a Geographic Environment System, for a solid visualisation and interpretation on a landscape level.

Fig. 7

(Left) 1968 aerial photograph: red dot on bottom left view indicates a specific area within the Eastern necropolis of Nea Paphos, with notable archaeological proxy concentration. A closer view of this area is shown in the background image. Yellow circles denote archaeological proxies. (Lysandrou & Agapiou, 2020); (Right) Arable areas and dense vegetation zones within the Eastern necropolis of Nea Paphos are shown in green colour. From the inspection of the area and concerning the GRVI values, a threshold was set to defined vegetated areas. (Lysandrou & Agapiou, 2020)

Important issues related to the risk management of monuments and archaeological sites in Cyprus are earthquakes, and urbanisation. Several studies that integrated satellite imagery and geo-spatial modelling have been carried out for the examination of the specific hazards. An example from a recent earthquake (2015) held in Paphos is reported by Agapiou and Lysandrou (2020). That study presented the results from the exploitation of a big-data cloud platform (Hybrid Pluggable Processing Pipeline-HyP3), for detecting ground displacement after a 5.6 magnitude scale earthquake in 2015. Ascending and descending pairs of Sentinel-1 images, acquired before and after the event, were processed through the HyP3 platform, revealing small relative ground displacements near the ‘Tombs of the Kings’ necropolis, and the Nea Paphos archaeological site (Fig. 8). As shown in Fig. 8, each estimated fringe corresponds to a change in range of λ/2, where λ is the Sentinel-1 radar wavelength (estimated to 5.54 cm for the Sentinel-1 radar satellite). The closer the fringes are together, the higher the deformation on the ground.

Fig. 8

(a) Unwrapped interferogram. (b) Vertical displacements. (c) Coherence map, enveloping important archaeological sites of Cyprus. (Agapiou & Lysandrou, 2020)

Urbanisation processes in Cyprus were documented through remote sensing sensors. For instance, in the Paphos District an increase of 300% of the urban footprint was mapped after analysing Landsat 5 TM and Landsat 7 ETM+ images. A supervised classification analysis covering the period 1980–2010 was examined by Agapiou et al. (2015). The expansion was recorded in the western part of the Paphos city; however, the rest of the island cities have shown a similar trend (Fig. 9). During this period, archaeological rescue excavations revealed significant archaeological records, like the Hellenistic and Roman tombs in Paphos (Lysandrou & Agapiou, 2020; Lysandrou et al., 2018), underlining the numerous subsurface wealth of archaeological findings of the island.

Fig. 9

Left: Urban expansion of the Paphos city from 1984 to 2010. Black colour indicate urban areas back in 1984; orange colour the urban areas of 1990; red colour the urban areas of 2000 and green colour the urban areas of 2010. Right: Land use change (%) from 1984 until 2010. Areas calculated from the Support Vector Machine (SVM) classifier and the Landsat images (top). Urban expansion (%) from 1984 until 2010. Areas calculated from the SVM classifier. During the last years, rescue excavations by the Department of Antiquities have been increased according to the yellow line shown in the graph (bottom)

Since then and after the economic crisis that hit Cyprus in 2012, the construction industry was considered a central pillar for the country’s future economic growth. Since 2015 land development was supported through large constructions and extended infrastructural works beyond human scale. This is a phenomenon known in the literature as vertical sprawl. Agapiou (2021) implemented a quick, automatic detection method using radar medium resolution Sentinel-1 images (Fig. 10). The approach was to capture the urbanisation process of the city of Limassol that has been initiated during this period due to recent large construction projects.

Fig. 10

Top: Radar polarisation visualisation change detection during the period 2015–2020 over Limassol city. Bright colours indicate areas of changes during this period. Bottom: The same are using as a background a high-resolution satellite image. (Source: ArcGIS Basemap)

6 Discussion and Final Remarks

The scope of this chapter was to summarise the contribution of various remote sensing sensors used in Cyprus over the last years, related to landscape archaeology and built heritage. Different prospection methods have been implemented through the years all over the island, spanning from ground-based techniques, aerial investigations, and satellite observations. Despite the number of projects that have been carried out, the employment of the specific techniques has been carried out in a non-systematic way and without specific planning, mostly based on the needs of the individual archaeological research campaigns. Furthermore, most of the programs that involve remote sensing techniques in the cultural landscapes have been realised at the southern part of the island, whereas almost none of them has been experienced in the northern Turkish occupied part of the island. Only the very early use of magnetometry at the Late Bronze Age site of Enkomi (Aitken, 1971) before the invasion has been recorded within the occupied part of Cyprus.

Most of the geophysical surveys have been carried out from foreign research teams, which most of the times were not aware of the soil or geological conditions of the area under study. The metamorphic geology and the rich in iron content soil deposits create limitations in some of the methods and this has been obvious at the results obtained. The lack of extensive soil analyses and the metadata (many of the surveys have not created extensive reports or have not been published) did not allow for the further enhancement of the methods. Early explorations were primarily aimed at recognising the extent of occupation of the sites rather than defining their detailed mapping. Advances in geophysical instrumentation now allow the detection of small structures, such as post holes, and can provide evidence for the presence or absence of archaeological structures, their depth and geometry with a much better resolution than in the past.

In a number of times, past human interventions have polluted the areas of them with modern debris which creates problems in the acquisition of quality geophysical measurements. In general, most of the geophysical surveys indicated increased levels of noise originating from the intense cultivation practices. At Idalion, the cultivation activities may have seriously affected the conservation of the ancient architecture in various parts of the site. Very few architectural structures remain in good preservation within the disturbed layers (Sarris, 2020). At Yeroskipou—Ayioi Pente, the area was severely disturbed due to modern human interventions in the area (agricultural activity, road and earthworks, mausoleum construction, etc.) (Sarris & Papadopoulos, 2010). Ancient architectural remains at Ayia Marina—Mavrovouni might have been dismantled by posterior clearance for agricultural purposes (Graham et al., 2013).

Similarly, the economic development on the island has followed an exponential growth since the 1980s. The emphasis on the tourist development, the large constructions all over the island, in the cities, villages and the coast, accompanied by the urban and population growth, easily depicted by the satellite imagery, had noticeable effects on the cultural archaeological sites and historical monuments (Agapiou et al., 2015). The intense cultivation, expansion of urbanisation and large construction works have been responsible for the destruction and bad preservation of archaeological remains. Taking into account the projection of the United Nations that about 69% of the population will be confined in urban centres by the year 2050, it is obvious that there will be an increased pressure for further expansion of the urban fabric and thus a higher level of threat in the enclosed and surrounding cultural monuments (Kiruthiga & Thirumaran, 2019) and this will also affect the cultural assets of the island.

Despite the few cores and soil analysis that has been conducted aiming towards the reconstruction of palaeoenvironmental conditions of the islands, such studies remain very limited. Most of the work has been conducted at the salt lakes of Larnaka and Lemesos (Akrotiri). The most recent studies deal with the palaeogeographic evolution of the closed lagoon of the Akrotiri Salt Lake based on the sedimentological and micropalaentological analyses of cores (Polidorou et al., 2021a) and the use of beachrock development as an index of the coastal changes in the past 2000 years (Polidorou et al., 2021b). From a much wider GIS and Remote sensing perspective, Agapiou et al. (2017) has considered the risk assessment of the coastal heritage landscapes within the marine spatial planning, but without taking into consideration the coastline evolution. Similarly, Andreou et al. (2017) and Andreou (2018) addressed the issue of the impact of the coastal erosion on the archaeological sites of Cyprus using the DSAS model for classifying the coastal erosion sections in a small section along the south-central part of the island (between Pyla and Tochni-Lakkia), with emphasis at Tochni-Lakkia and using a combination of historical aerial photos of the Department of Lands and Surveys, laser scanning and geophysical techniques.

These past advances indicate the accelerated momentum that has been achieved in the application of remote sensing and GIS techniques in the archaeological context of Cyprus. This has driven to some latest developments that are expected to have a major impact in the research and Cultural Heritage management (CHM) of the island. The first concerns the establishment of devoted laboratory research units at the University of Cyprus (UCy) and the Cyprus University of Technology (CUT). At the University of Cyprus, the Laboratory of Digital Humanities GeoInformatics (DigHumanities GeoInfo Lab—http://www.ucy.ac.cy/geoinfolab/the-lab), supported by the “Sylvia Ioannou” Chair for Digital Humanities at the Department of History and Archaeology is dealing with the application of cutting-edge technologies (Geographical Information Systems (GIS), Geoinformatics, computational and statistical methods, etc.) in Landscape Archaeology, Environmental Archaeology, Monument Monitoring, and Cultural Heritage Risk Assessment and Modelling. The Lab has already participated in a number of projects in Cyprus (Palaepaphos, Xeros valley, Fraggisa, Idalion, Paralimni, Petounda), Greece (Pylos, Thouria, Rethymno, ancient Corinth, ancient Halos) and Kosovo (RAPID project), emphasising its research in the Eastern Mediterranean and the Balkans. At the same time, a new graduate program on Digital Heritage and Landscape Archaeology (http://www.ucy.ac.cy/mscgidh/) has been initiated at the University of Cyprus aiming to provide training to international students on the application of Spatial Technologies and GeoInformatics in the wider domain of Digital Humanities. Similar is the case of the former Archaeology and Cultural Heritage section (ARCH) of the Remote Sensing and Geo-environment research Lab (http://www.cyprusremotesensing.com/) at the Cyprus University of Technology, the research of which emphasises the 2D/3D documentation of archaeological sites, the aerial and satellite remote sensing applications, the GIS management of cultural heritage sites, and geoinformatics for archaeological analysis (Archaeology and Cultural Heritage section, 2021). The group has been active in a number of research projects in Cyprus (Kataliondas-Kourvellos, Pano and Kato Pyrgos Tillirias, Mansoura Tillirias, Ayios Sozomenos, Palaepaphos, Nea Paphos, Politiko, Sotira, Paralimni, Xeros valley) and Greece (Malia). Also, is organising several training courses and workshops. The group has made clear that in Cyprus, integrated remote sensing state-of-the-art techniques have been implemented, also following the recent trends of related to the Copernicus European Space Programme with the Sentinel missions, as well as the exploitation of big data and earth cloud platforms.

It is obvious that though the collaboration of the above groups, as this joint paper indicates, there will be more systematic engagements expected in the domain of remote sensing techniques in archaeology, which is taken positively by the Department of Antiquities of Cyprus, as it will be able to exploit the above resources in the best possible way. Furthermore, the above action will improve the scientific capacity to support the various needs from the archaeological research and heritage management perspectives, creating a roadmap for the more systematic use of these technologies by the local and foreign academic and research institutions.