Post-print version of :
Orengo, H.A.; Krahtopoulou, A.; Garcia-Molsosa, A.; Palaiochoritis, K. & Stamati, A. 2015. Journal of
Archaeological Science, 64: 100-109.
http://dx.doi.org/10.1016/j.jas.2015.10.008

Photogrammetric re-discovery of the hidden long-term landscapes of western
Thessaly, central Greece
a, *

, A. Krahtopoulou b, A. Garcia-Molsosa c, K. Palaiochoritis b, A. Stamati

H.A. Orengo
a
b
c

b

Department of Archaeology, University of Sheffield, Northgate House, West Street, Sheffield, S1 4ET, United Kingdom
Ephorate of Antiquities of Karditsa, 1 Loukianou Street, Karditsa, 43100, Greece
Catalan Institute of Classical Archaeology, Pl. Rovellat s/n, Tarragona, 43003, Spain

a r t i c l e

i n f o

Article history:
Received 25 May 2015
Received in revised form 20 September 2015
Accepted 15 October 2015
Available online 31 October 2015

Keywords:
Remote sensing
Photogrammetry
GIS
Ancient landscape reconstruction
Thessaly
Central Greece

a b s t r a c t
This paper introduces a novel workflow for the reconstruction of nowadays disappeared cultural landscapes based
on the extraction of morphological information from historic aerial photographs. This methodology has been
applied for the first time for the detection, classification and characterisation of upstanding, flattened and buried
archaeological sites and various off-site ancient landscape features in the plain of Karditsa, western Thessaly.
Although Thessaly has been the focus of prehistoric, and especially Neolithic, research in Greece, since the
beginning of the 20th century, western Thessaly has not received as much archaeological attention and its
archaeological record remains rather scanty. Moreover, an extensive land reclamation project implemented in the
western Thessalian plain during the early 1970s resulted in the flattening of habitation tells and funerary sites of
all periods. Thus, recognition of archaeological sites and relict landscape features becomes extremely difficult,
whereas standard landscape analysis and application of mainstream Remote Sensing (RS) techniques based on
multispectral satellite images are problematic.
Digital photogrammetric reconstruction techniques and the subsequent GIS-based treatment of the results
allowed overcoming these challenging limitations: the combined use of pre-1970s aerial pho-tographs with later
imagery provided a powerful means to reconstruct the landscape before the land reclamation process, using a
workflow designed to highlight photogrammetry-derived topographic differences and multi-temporal imagery
analysis.
Hundreds of previously unknown mounded archaeological sites, as well as other ancient landscape traits such
as roads, city grids and field systems were detected. More importantly, invaluable insights into the type and
character of these archaeological features were gained, which would have been impossible to obtain by
conventional RS techniques.

1. Introduction
The pioneering efforts of Tsountas (1908), Wace and Thompson
(1912) and Mylonas (1928), and later seminal work by Miloj.ciIc
(1959), Theocharis (1967, 1973), Chourmouziadis (1971, 1979),
Gallis (1992), Kotsakis (1983) and Halstead (1981, 1984) have
brought Thessaly to the fore in Neolithic research in Greece
(Andreou et al., 1996) and have set the agenda for new excavations,
surface surveys, analysis of old excavation data and rigorous
theoretical discourse.

* Corresponding author.

E-mail address: Hector.Orengo@sheffield.ac.uk (H.A. Orengo).

Thessaly forms a closed geomorphological environment, bounded by mountain ranges and drained by the river Pineios and its
tributaries. To the interior, Middle Pleistocene tectonism created two
large basins, the Larissa to the east and the Karditsa-Trikala to the
west (Higgins and Higgins, 1996) ( Fig. 1). These basins were
subsequently filled with alluvial deposits (Demitrack, 1986) and today
they form the largest and among the most productive agricultural
lands in Greece.
Archaeological and geomorphological research focused on
eastern Thessaly, whereas western Thessaly received much
less archaeological and palaeoenvironmental attention (see
Andreou et al., 1996 for additional references). This is clearly
demonstrated by the IGEAN (2015) Neolithic archaeological site
inventory that reports 342 documented Neolithic tell-like sites, of

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

101

73-75). This project introduces a new methodological approach,
based on the digital reconstruction of the cultural landscape as it was
before the land reclamation process. This novel methodology
attempts to overcome the problems posed by the flattening of the
landscape in this reclaimed area, which renders impossible the
application of previously employed RS methods. Moreover, moving
beyond the detection of sites and monuments, this approach provides a more secure identification and a better characterisation of
the entire cultural record. This approach is considered of vital
importance, since the ultimate goal of this project is to understand
and reconstruct the long-term cultural landscapes of western
Thessaly.
2. Methodology and sources
2.1. The assemblage of the project's GIS geodatabase

Fig. 1. Location of the study area.

which 288 are located in eastern and 55 in western Thessaly. It is
often thought that the striking scarcity of prehistoric presence in
Western Thessaly is due to the “hostile” physiography of the area
(Andreou et al., 1996: 276). However, lack of intense surface
surveys, geoarchaeological and palaeoenvironmental work and final
publication of excavation and extensive archaeological reconnaissance results prevent any conclusive argument as to the
intensity and character of prehistoric habitation, as well as the nature
of land-scape and use of space through time.
In the heart of the Karditsa alluvial plain lies the former municipality (now municipal unit) of Kambos that occupies an area of
90 km2 (Fig. 1). Between 1970 and 1972, rescue excavations
were carried out by G. Chourmouziadis at three Neolithic sites
(Prodromos I, II and III) (Chourmouziadis, 1971, 1972). Very few
prehistoric and historical sites supplement the arguably poor
and sketchy archaeological map of the Kambos area (e.g. Intzesiloglou, 1997, 2000, 2010; Nikolaou, 2004; Chatziaggelakis,
2007; Chatziaggelakis and Karagiannopoulos, 2012).
To further complicate this very tentative picture, a large-scale,
centrally planned and executed project to reclaim land for intensive,
mechanised, irrigated agriculture was implemented in the Karditsa
plain during the early 1970s (from 1970 to 1972). As a result, the
landscape was dramatically trans-formed (Krahtopoulou et al. in
preparation) ( Fig. 2): the flow of major rivers and smaller streams
was modified to avoid seasonal flooding, wetlands were drained, an
extensive network of irrigation canals was installed, old roads and
fields were reshaped, new fields and road networks were created,
the land was flattened and soil was locally shifted to fill topographic
depressions. Moreover, mounded tell-like habitation sites, locally
known as “magoules”, characteristic of the Thessalian lowlands, as
well as other types of monuments, mainly tombs, were partially or
totally destroyed (Gallis, 1979: 18). Recent low intensity, extensive
landscape reconnaissance and informal interviews with members of
the local communities indicated that the cultural record of the
Kambos area was gravely affected by the land reclamation process
and especially by the establishment of a specific irrigation system
that required the systematic flattening of the landscape
(Krahtopoulou et al. in preparation).
Recent research has employed multispectral satellite imaging in
combination with geophysical survey and GIS to locate unexcavated
magoula sites in Thessaly with very promising results (e.g. Alexakis
et al., 2011; Agapiou et al., 2012, 2013, 2014; Hadjimitsis et al., 2013

Photo-cartographic material, necessary for the development of
this research, was provided by the Hellenic Military Geographical
Service (HMGS), the Hellenic Cadastre (Ktimatologio) and the
United States Geological Survey (USGS). The HMGS provided the
aerial photographs on photographic paper and they were scanned
with an A3 photogrammetric scanner at 900 dpi. The calibration
parameters of the cameras employed in the pre-1972 flights were
provided by the HGMS.
The first sources to be incorporated in the GIS geodatabase were
the reference cartographic materials. These were relatively easy to
incorporate by a simple georeferencing process in which the coordinate grids of the map series employed, provided in a local coordinate system, had to be reprojected to the Greek Grid (EPSG
2100) used in the development of the project. QGIS Georeferencer
Plugin was employed to do this. The sources integrated in the GIS
geodatabase were:
- 1963-73 cadastral maps at a scale of 1:5000. They afford high
accuracy and large scale. They also provide important information on the second- and third-level trigonometric network.
Interestingly enough, they sometimes depict elements of the
landscape with topographic relief. Due to the flat morphology of
the plain, magoules and tombs often provided ideal places for the
construction of these trigonometric points (Fig. 3).
- 1973-75 topographic maps at a scale of 1:5000 produced by the
HMGS. These were extremely useful since they contain detailed
topographic data including thousands of spot heights that were
employed to analyse post-reclamation topographic changes.
They also contain many post-1972 trigonometric points (most of
those appearing in the pre-1971 maps were destroyed together
with the magoula or tomb upon which they were positioned)
that were very useful as they could be used as Ground Control
Points (GCP) for the photogrammetric ortho-correction and
georeferencing of the later aerial photograph series.
Digital photogrammetric block triangulation was considered
necessary for the integration of the different series of photographic
materials. Two types of outputs were produced using this tech-nique:
(1) ortho-photomosaics of the study area for as many dates as
pre-1972 (the date by which the land reclamation process was
completed) photographic series were available and (2) Digital
Surface Models (DSM) of the area. The development of the orthophotomosaics and DSMs followed the same process of photogrammetric correlation between the different images of each flight.
To assist this process and assure the correct georeferencing of the
products, a set of accurate GCPs were acquired using the previously
discussed projected cartographic material and identified in each
image.

102

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

Fig. 2. Photograph of the land reclamation process, which flattened the area under study effectively destroying most archaeological remains (Administrative Region of Thessaly, photographic archive).

Fig. 3. Georeferenced scanned map (red lines) on top of the 1960 ortho-photomosaic. Towards the center, a trigonometric point (red triangle) rests on top of a nowadays invisible magoula (Мagούla is the
toponym).

The different photographic materials of interest for the study
area incorporated into the project's GIS geodatabase were:
- The 1945 flight at a scale of 1:42,000 acquired by the Royal Air
Force (RAF). Although these photographs could be employed for
the visual detection of features, the scale was too small to allow
the use of digital photogrammetric techniques for the reconstruction of the area's topography at a significant resolution.
- Two 1960 flights at a scale of 1:15,000 and 1:30,000 acquired by
the U.S. Air Force (USAF). The 1: 30,000 photographs offer a
good complement to the 1:15,000. Although these cannot be
used to produce useful DSMs due to their scale, they can be
used for the visual detection of features and increase the
number of images available, which is particularly important given
the problems of

image contrast apparent in the 1:15,000 images (see Fig. 4). The
1960 flight at a scale of 1:15,000 allowed the creation of both an
ortho-photomosaic (Fig. 4) and a relatively high resolution DSM.
Although these images provided a reasonably high resolution at
an early enough date, the evident problems of contrast hindered
in some cases the development of DSMs, which showed inconsistencies such as pits or peaks. However, the resulting
DSMs still allowed us to reconstruct the topography of now
invisible magoules and tombs.
- CORONA analogue satellite imagery was also employed. In the
Kambos case CORONA offers excellent image quality and
a ground resolution of 2 m. Two strips of CORONA coverage
acquired in Jun 1970 were obtained from the USGS, georeferenced, employing the high resolution GCPs, and incorporated into the

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

103

sites, most of them still upstanding. Although this is a rather
modest record, it provides valuable chronological and
morphological information that can be related to different types of
features but also criteria for the identification of micro-reliefs that
have been related to former archaeological sites.
- The first-order trigonometric network of the area was also
incorporated into the GIS as a point layer depicting the position
of current trigonometric vertices in the area.
2.2. Methodological approaches for the reconstruction of
Thessalian hidden landscapes
In order to compensate for the reduction in spectral information
with respect to modern multispectral imagery that the use of old
monochrome aerials would imply, it was decided to increase the
chances of feature detection by employing a workflow (Fig. 5)
specifically designed to extract the maximum information possible
from old black-and-white aerial photographs. The workflow was
based on two methodological approaches:
- Photogrammetric 3D reconstruction. Moving from the sole visual interpretation of features and taking advantage of the
stereo-photogrammetric capabilities that old near-vertical
aerial photographs offer, digital photogrammetric block triangulation was used for the 3D reconstruction of the landscape as
it was before the 1970s. Although this is not a new development,
few archaeological applications take advantage of the topographical information enclosed in old aerial and/or satellite
imagery stereopairs for the production of DSMs (but see the
innovative work of Casana and Cothren, 2008 and Galiatsatos
et al., 2008 with CORONA satellite images and, more recently
Orengo et al., 2014 with old aerials).

Fig. 4. Ortho-photomosaic developed from the 1960 aerial photographs of the study area. The
Kambos sub-municipal unit is delineated in red.

project's GIS. At the date when the CORONA images were taken
the land reclamation process was still in progress and some
areas, such as the eastern sector of the study area, had still not
been reclaimed. Even in those sectors where the landreclamation process was ongoing some magoules, tombs and
older field systems were still visible.
- The digital orthoimages of the Hellenic Cadastre (Ktimatologio)
aerial flight (2007e9) were incorporated as a reference base-map.
These images reflect the landscape as it is today and allow us to
check whether any of the sites located through remote sensing
techniques are still standing.
Lastly, two point-vector layers providing important information
were incorporated into the project's geodatabase:
- The sites' record database of the Ephorate of Antiquities of
Karditsa was integrated into the geodatabase. This database
incorporates sites reported in the relevant bibliography, as well as
sites discovered by recent extensive reconnaissance of the area
(Krahtopoulou et al., in preparation). The coordinate fields X and
Y were used to produce a point vector layer from which all
information related to the sites could be queried and analysed.
The layer attribute table contained information on site name,
area, chronology, type of archaeological exploration, if any, and
preservation. The layer presents 36 habitation and funerary

The analysis of topographic change in the area employed the
DSMs developed from both pre- and post-1970 aerials. In order to
obtain a fast and reliable picture of topographic changes, these two
models were compared using a cut-and-fill process. The models did
not completely overlap on the Z axis. Small changes in the DSM
values due to the interpolation process and the differential availability of GCPs for the different periods were probably responsible
and, although changes were still appreciable, these were not as
clear and uniform as expected. In order to address this problem in
the comparison of models, we used an algorithm developed for the
filtering and visualisation of LiDAR data: Local Relief Model (LRM)
(Hesse, 2010). LRM produces a raster in which only relative topographic differences are reflected and, in doing so, reducing constant
trends in height changes to a flat surface. This technique is particularly appropriate for the comparison of DSMs, since the cut-and-fill
between the pre- and post-1970 LRMs will find no difference in the
areas where topography is constant in both, and only when there are
significant differences in relative (not absolute) height will these be
picked up by the cut-and-fill tool. Although some of those differences
in height highlighted by the cut-and-fill process were due to pits and
peaks generated by the photogrammetric reconstruction process,
easily detectable by their morphology and the comparison with the
ortho-photomosaic, many corresponded to flattened out mounded
sites. The sites located could be further investigated by using the
original pre-1970 DSM where absolute measurements could be
obtained and their shape analysed.
This novel workflow combining photogrammetric-generated
DSMs, LRM and cut-and-fill DSM comparison allows the fast and
accurate analysis of recent topographic change, an important capacity since most large-scale anthropogenic topographic changes in
the world have taken place during the last 60 years due to urban

104

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

Fig. 5. Scheme of the workflow followed for the photogrammetric-based multi-temporal Remote Sensing (RS) analysis of the study area.

development, the global expanse of mechanised agricultural practises and the use of heavy machinery. In this way it would be
possible to detect features flattened out during the land reclama-tion
process, such as magoula settlements or tombs, through their
characteristic topography.

- Multi-temporal visualisation. The visual detection of features is
strongly linked to physical differences in ground conditions,
which, at the same time, are bound to sometimes rapidly
changing factors, such as time of acquisition, soil moisture,
temperature, and so forth. Moreover, multi-temporal data were

Fig. 6. Multitemporal RGB image generated from the 3 Principal Components of a PCA analysis of the different orthoimages available for the study area.

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

105

Fig. 7. Examples of mounded sites (yellow lines) detected through topographic reconstruction and RS detected roads (red lines) in the area east of the modern village of Agios Theodoros. Some of those
sites identified only as topographic marks are enlarged in the images in the boxes. None of these exist today as shown in the Ktimatologio orthopho-tographs below.

used to increase the chances of feature detection. The photogrammetric 3D reconstruction had a second output besides
DSMs: ortho-photomosaics of the different flights employed were
produced and then combined in a single multi-temporal image.
This image was created using a Principal Component Analysis
(PCA), a method available in most GIS software packages, of all
ortho-photomosaics. The three principal components were
combined in a RGB (Red-Green-Blue) composite (Fig. 6). This
composite provided a means to explore the diversity of the data
in a simple and efficient way, taking into account variation in
environmental conditions, but also in human transformations,
which could have caused differential visibility of features.

2.3. RS-based interpretation process
The multi-temporal image together with the 1945 and 1960 orthophotomosaics provided an excellent tool for the visual inspection of
the study area. Their high resolution of 1 m/px, 1.55 m/px and 0.657
m/px respectively, was high enough to allow visibility of the smallest
features. Image contrast, however, was in some cases problematic,
hindering the identification of many features that had to be identified
by using alternative sources to the initial visual inspection.
The analysis of the multi-temporal image followed a systematic
approach: a vector grid in which each grid cell was equal to the size
of the map window at the 1960s ortho-photomosaic's maximum

resolution was created and laid over the study area. To secure a
systematic inspection each grid square was visually investigated
separately. This operation was repeated for each ortho-photomosaic
and the results compared. Although most soil marks were visible in
several images, some were only identifiable in one of them,
highlighting the importance of employing images acquired in different
environmental conditions to increase the chances of feature
detection.
Archaeomorphological analysis was also performed following the
methodology developed by Orengo and Palet (2010) and Palet and
Orengo (2011). Archaeomorphology, a term employed mainly in
French landscape studies, refers to the study of ancient humanmade structures preserved as part of the modern landscape, such as
roads, field systems, channels and irrigation systems, and urban
shape and distribution by means of cartographical documents and
aerial photography. In the Kambos case, the results were significant
as they allowed the ordering of feature morphologies, such as roads,
into a relative chronological sequence. Also, road morphology was
an important factor in ascribing typologies to mounded sites and/or
features according to the location of the site/feature in the road
network and its relation to other sites/features inside the network.
The photointerpretation was made in a polyline vector layer. The
attribute table linked to this layer incorporated information of the
type of feature detected (for example, road, mounded site, wetland,
etc.), subtype (for example, different types of mounded sites or
roads were given different codes), chronology, when available, the

106

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

in LiDAR datasets. These included Local Relief Model (Hesse,
2010), which, as already mentioned, was incorporated as part of the
cut-and-fill production process, and a RGB composite of the 3 principal components of a PCA of multiple shaded relief models
(following Devereux et al., 2008). The results not only allowed us to
locate mounded sites but, in many cases, to map their topography.
3. Results and discussion
Four types of structures, according to the factors prompting
their detection, have been identified:
- Sunken or hollow structures. The most common structures of
this type are the so-called “hollow ways” (Wilkinson, 1993). The
continuous use of these informal paths by people, animals and
vehicles erodes the substrate and forms a depression. The geology of the plain of Karditsa, namely easily erodible Quaternary
alluvial deposits, promotes the formation and preservation of
such negative relief features. Moreover, the plain forms a large
natural basin that retains water perennially. The water table is
particularly high and oscillates between 2 and 4 m. Therefore
negative relief features, such as hollow ways, ditches, and so on
are unusually visible, as they tend to present different moisture
content than that of their immediate environment.
Marks related to differential vegetation growth such as formal
- �
paved roads, but also ancient city grids and walls.
Preserved and standing structures (at the moment at which the
image was acquired). These are in the majority magoula settlements or funerary sites, which are usually visible in aerial photographs but also detectable through the photogrammetric 3D
models developed (Fig. 7). This category also includes structures
and landforms located through archaeomorphological analyses,
such as field systems or ancient rivers. These last are usually
detected using visual identification and a series of morphological
factors such as the existence of constant distances between field
limits related to ancient measure units, interrelation between
structures, or characteristic morphologies.
Structures located through old documents, cartographic analysis
and ethnographic work. Fig. 3 offers a good example of one of
those.

Fig. 8. Results of the photointerpretation and archaeo-morphological analysis.

type of source employed for its detection, the year in which this
source was created and finally the degree of security of the identification. This last information was based on the consistency of site/
feature detection. Elements appearing in maps, such as the one
shown in Fig. 3 had a high level of reliability, as did those appearing
in at least two images and, at the same time, showing characteristic
topography in the DSMs. Elements appearing in more than one
source had a medium level of reliability, while those appearing in
only one source were tagged as low level.
The DSM generated from the 1960 photographs presented a
spatial resolution of 1.329 m/cell but, due to the average quality of
the images, it presented several deformations in specific areas.
These were treated with standard tools but the DSM was not filtered
in any way, in order to preserve small depressions, which could
reflect actual data. The visual inspection of the DSM made use of
different filtering and visualisation methods, from the simple
histogram stretch of the height values to more sophisticated
techniques initially developed for the visualisation of micro-reliefs

In total, 891 previously unknown features of archaeological interest were detected. These can be broadly classified as mounded
or rounded features (377), road stretches (446), elements of field
systems (22) and isolated finds of diverse types (46) (see Fig. 8 for a
general map of detected features). This paper focuses on mounded
sites and roads, leaving the interpretation of the communication
network and the analysis of the area's historical morphology for a
second phase of work in which the development of archaeological
survey campaigns will provide data to inform the historical interpretation of the detected features.
Our results prove that the Kambos area preserves a much richer,
denser and more diverse network of magoules and tombs than
previously thought. It is indicative that the IGEAN (2015)
archaeological site inventory reported fifty documented Neolithic
magoules in the Karditsa prefecture. Only two of those are located
in the Kambos sub-municipal unit. Preliminary identification of 377
unspecified mounded sites is thus a sub-stantial improvement of our
knowledge of the cultural record of the area.
Comparison of the analysis results with the modern Ktimatolo-gio
orthoimages and available archaeological site inventory showed that
very few of the magoules and tombs detected are presently
preserved. In several cases the DSMs were able to reconstruct the
topography of invisible or partly-flattened features (Figs. 7 and 9).

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

107

Fig. 9. Topographic reconstruction of the Late Neolithic/Bronze Age Karnomagoula south of the modern village of Pteloupula (zoom-image of the magoula in the bottom right image).
Photointerpretation from the 1960s aerials (superimposed on the Ktimatologio orthophotograph in the top left image).

They provide therefore a highly reliable counterpart to the visual
identification provided by the multi-temporal image and the 1945 and
the 1960 aerial series ortho-photomosaics. More importantly the 3D
reconstruction of the topography of the study area as it was in 1960
allowed the detection of magoules and tombs not identified by the
aerial photographs' visual photointerpretation process. Initial ground
checks seem to verify the results of the landscape analysis. Only
nineteen of the flattened sites detected through the analysis have
been visited so far (Krahtopoulou et al., in preparation). They all
preserve surface archaeological material, thus confirming the huge
potential of this methodology.
Our results have gone beyond the simple identification of
magoules and tombs. The combination of the detected archaeological features with the results of archaeomorphological analysis
and the available archaeological data provided meaningful insights
into the character of mounded sites and their chrono-typological
ascription. Three main sub-typological groups were detected. The
first group includes large magoules, in most cases larger than 30 m
in diameter (the larger examples reaching 90 m in diameter), with
rounded morphology. These are interpreted as habitation sites.
Archaeomorphological analysis shows that some of these are
located at the centre of radial road networks. The Late NeolithicBronze Age site of Karnomagoula shown in Fig. 9 offers a good
example of this type of settlement. The few additional examples for
which dates are presently available point to a similar chronological
framework. The second group involves large magoula settlements
that are joined by a well-developed network of relatively straight
roads, perhaps indicating simultaneous occupation. The few

presently dated examples imply a historical time framework
(Classical to Roman period).
The last sub-group comprises smaller mounded features (5e20 m
in diameter with a few larger than 30 m) that lie along roads, are not
surrounded by radial hollow way systems and sometimes exhibit
small depressions around them. These are interpreted as funerary
sites. Available archaeological data from the Kambos area show that
tombs of this type have been built from at least the Proto-Geometric
to the Roman period (from the 10the8th century BCE to the 4th
century AD). In Pieria, Macedonia, northern Greece, cemeteries
dating from the Early Bronze Age (3300/3100-2300/2100 BCE)
onwards, and isolated tombs, dating to the late Classical-early
Hellenistic (4th century BCE) period, were also located along roads
(Besios, 2003, 2010; Besios and Athanasiadou, 2014; Besios and
Krahtopoulou, 2003; Besios et al., 2003). In the Kambos case,
association with excavated examples (e.g. Intzesiloglou, 1995, 2000,
2010) indicates that tombs were placed along roads since at least
the Proto-Geometric period (10th-8th century BCE). Some of the
tombs detected are surrounded by an excavated ditch (Fig. 4, tombs
2, 3 and 4; tomb 1 and, to some extent, tomb 3 show small
depressions around). The erection of tombs involved digging up soil
around them (Besios, unpublished data). The soil was then piled on
top to create the impressive monuments that once littered the
nowadays flat and featureless Kambos landscape.
This project has also developed sub-typologies of roads, namely
paved roads, hollow ways and double ditched roads. These buried
features were particularly visible due to the combination of their

108

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

morphology and the geological and hydrological characteristics of
the Karditsa plain. A buried formal paved road identified by this
project measures up 4012 m in length and leads to the ClassicalHellenistic-Roman site of Kierion. An excavated segment of this road
verifies our interpretation and indicates that paved roads were in use
at least since the 2nd-1st century BCE, during the Hellenistic period
(Intzesiloglou, 2010). Only one example of a double ditched road has
been found in the area. It runs in straight stretches for 8700 m,
crossing the northern sector of the study area following a N-NW
direction. Although no secure chronological ascription can be
assumed at the moment, its relation to field systems with a distance
between field limits corresponding to multiples of historical measure
units, which seem to be aligned to it, might offer a preliminary dating.
Negative relief features known as ‘sunken lanes’ or ‘hollow
ways’ (Wilkinson, 1993) are a common indication of ancient informal
road systems. Linear and radial hollow way networks are widespread
and exceptionally well documented at a variety of scales in the
Middle East (e.g., Casana, 2013; Menze and Ur, 2012; Wilkinson et
al., 2010). In Pieria, northern Greece excavated linear examples
date from the Early Bronze Age to the 14th century AD (e.g. Besios,
2003, 2010; Besios and Athanasiadou, 2014; Besios and
Krahtopoulou, 2003; Besios et al., 2003). Pedogenic expression of
the fill of a currently excavated hollow way in the Kambos area
suggests that this pathway was used during prehistoric times.
Completion of the excavation will provide a more secure and close
chronology.
4. Conclusions and future work
This paper has highlighted the enormous potential of the
application of multiple remote sensing techniques and sources for
the analysis, reconstruction and understanding of lost cultural
landscapes. The photogrammetric reconstruction of the Kambos
area succeeded in locating hundreds of no longer visible or
compromised archaeological features dating from the Early Neolithic
onwards, clearly proving that existing knowledge on the cultural
record of the area is spatially inconsistent, incomplete and
misleading. The use of old aerial photographs provided a unique way
to undo virtually the ravages of the last fifty years, when land
reclamation, urban development and infrastructure projects led to
large-scale and often irreversible landscape modifications. This
technique was combined with more traditional visual photointerpretation of the ortho-photomosaics generated during the same
photogrammetric reconstruction process. Archaeomorphological
analysis offered invaluable new perspectives on the nature of the
detected structures, while integration with available archaeological
information provided meaningful chronological associations.
Initial targeted surface archaeological work, albeit limited, appears to validate our results. Ground checks will continue rigorously,
in order to gather more data and to refine, if necessary, classification
and characterisation of the archaeological features detected. Dating
of the surface material will establish firm chro-nologies for most
identified features. Chronologies will be then securely related to the
different morphological types identified during this initial phase of the
project. Moreover, trial trenches will provide ground truthing, formation and dating evidence for hollow ways and paved roads
recognised by this project. Finally, ongoing geoarchaeological work
is expected to elucidate the role of geomorphological processes,
mainly alluviation, in the detection potential of the methodology
introduced in this paper.
The data gathered by this project have been made available to
the local Archaeological Service and they already constitute a
unique resource for the protection of the entire cultural heritage,

upstanding, buried and otherwise, and for the implementation of
archaeological legislation when necessary.
Acknowledgements
First and foremost our thanks to the Institute for Aegean Prehistory (INSTAP) for providing funds without which this research
could have not been conducted. The Nottingham Advance Research
Fellowship scheme of the University of Nottingham and the Marie
Curie Intra-European Fellowship provided research funding for
Hector Orengo. We would like to thank the Administrative Region
of Thessaly for allowing access to the land reclamation project archives. Our gratitude also to Prof. John Bennet, Prof. Andrew Bevan
and Prof. Paul Halstead who read a preliminary version of this
paper and provided useful comments and valuable advice. Two
anonymous referees provided constructive comments and useful
corrections. As usual, all errors remain our own.
References
Agapiou, A., Alexakis, D.D., Sarris, A., Hadjimitsis, D.G., 2013. Orthogonal Equations of MultiSpectral Satellite Imagery for the identification of Un-Excavated Archaeological Sites. Remote
Sensing, 5, 6560-6586.
Agapiou, A., Hadjimitsis, D.G., Alexakis, D.D., Sarris, A., 2012. Observatory validation of Neolithic
tells (“Magoules”) in the Thessalian plain, central Greece, using hyperspectral spectroradiometric
data.
Journal
of
Archaeological
Science,
39,
1499-1512.
Agapiou, A., Alexakis, D.D., Sarris, A., Hadjimitsis, D.G., 2014. Evaluating the Potentials of
Sentinel-2 for Archaeological Perspective. Remote Sensing, 6, 2176-2194.
Alexakis, D., Sarris, A., Astaras, T., Albanakis, K., 2011. Integrated GIS, remote sensing and
geomorphologic approaches for the reconstruction of the landscape habitation of Thessaly during
the neolithic period. Journal of Archaeological Science, 38(1), 89-100.
Andreou, S., Fotiadis, M., Kotsakis, K., 1996. Review of Aegean Prehistory V: the Neolithic and
Bronze Age of Northern Greece. American Journal of Archaeology, 100, 537-597.
Besios, M., 2003. Νότιο νεκροταφείο Πύδνας. Το Αρχαιολογικό Έργο στη Μακεδονία και Θράκη,
15(2001), 369-377.
Besios, M., 2010. Πιερίδων Στέφανος: Πύδνα, Μεθώνη και οι αρχαιότητες της βόρειας Πιερίας.
ΑΦΕ, Κατερίνη.
Besios, M., Athanasiadou, A., 2014. Νεκροταφεία Πύδνας: ανασκαφή στο αγροτεμάχιο αριθ. 691
Μακρυγιάλου. Το Αρχαιολογικό Έργο στη Μακεδονία και Θράκη, 24(2010), 127-134.
Besios, M., Athanasiadou, A., Gourtzioumi, I., Karanikou, Z., Noulas, K., Christakou-Tolia, M.,
2003. Ανασκαφές βόρειας Πιερίας. Το Αρχαιολογικό Έργο στη Μακεδονία και Θράκη, 15(2001),
379-384.
Besios, M., Krahtopoulou, A., 2003. Η εξέλιξη του τοπίου στη βόρεια Πιερία. Το Αρχαιολογικό
Έργο στη Μακεδονία και Θράκη, 15(2001), 385-400.
Casana, J., 2013. Radial route systems and agro-pastoral strategies in the Fertile Crescent: new
discoveries from western Syria and southwestern Iran. Journal of Anthropological Archaeology,
32(2), 257-273.
Casana, J., Cothren, J., 2008. Stereo analysis, DEM extraction and orthorectification of CORONA
satellite imagery: archaeological applications from the Near East. Antiquity, 82(317), 732-749.
Chatziaggelakis, L.P., 2007. Προϊστορικοί και Ιστορικοί Χρόνοι. In Tsagaraki, E., (ed), Οδοιπορικό
στα Μνημεία του Νομού Καρδίτσας: 15-82. Νομαρχιακή Αυτοδιοίκηση Καρδίτσας, Καρδίτσα.
Chatziaggelakis, L.P., Karagiannopoulos, Ch., 2012. Πρόδρομος Καρδίτσας. Νεότερα στοιχεία
από τη Μαγούλα στον Άγιο Ιωάννη. ΑΕΣΘΕ, 3(2009), 85-96.
Chourmouziadis, G. 1971. Δύο νέαι εγκταστάσεις της αρχαιοτέρας Νεολιθικής εις την Δυτικήν
Θεσσαλίαν. Αρχαιολογικά Ανάλεκτα εξ Αθηνών, 4(2), 164-175.
Chourmouziadis, G. 1972. Ανασκαφές εις τον Πρόδρομον Καρδίτσης. Αρχαιολογικό Δελτίο, 27,
394-396.
Chourmouziadis, G. 1979. Το Νεολιθικό Διμήνι. Βόλος, Εταιρεία Θεσσαλικών Ερευνών.
Demitrack, A. 1986. The Late Quaternary Geological History of the Larissa Plain, Thessaly,
Greece: Tectonic, Climatic and Human Impact on the Landscape. Unpublished Doctoral
Dissertation, Stanford University.
Devereux, B.J., Amable, G.S., Crow, P., 2008. Visualisation of LiDAR terrain models for
archaeological feature detection. Antiquity, 82, 470-479.
Galiatsatos, N., Donoghue, D.N.M., Philip, G. 2008. High Resolution Elevation Data Derived from
Stereoscopic CORONA Imagery with Minimal Ground Control: An Approach Using Ikonos and
SRTM Data. Photogrammetric Engineering & Remote Sensing, 74(9), 1093-1106.
Gallis, K. 1992. Άτλας Προϊστορικών Οικισμών της Ανατολικής Θεσσαλικής Πεδιάδας. Εταιρεία
Ιστορικών Ερευνών Θεσσαλίας, Λάρισα.
Gallis, C.J., 1979. A short Chronicle of the Greek Archaeological Investigations in Thessaly, from
1881 until to the present day. In Pouilloux, J. (Ed.), La Thessalie. Actes de la Table-Ronde, 21-24
juillet 1975. Maison de l'Orient et de la Méditerranée, Lyon, pp. 1-30.
Halstead, P., 1981. Counting sheep in Neolithic and Bronze Age Greece, in: Hodder, I., Isaac, G.,
Hammond, N., (Eds.), Patterns of the Past. Studies in Honour of David Clarke. Cambridge
University Press, Cambridge, pp. 307-339.

H.A. Orengo et al. / Journal of Archaeological Science 64 (2015) 100-109

Halstead, P., 1984. Strategies for Survival: An Ecological Approach to Social and Economic
Change in the Early Farming Communities of Thessaly, Northern Greece. Unpublished Doctoral
Dissertation, University of Cambridge.
Hadjimitsis, D.G., Agapiou, A., Themistocleous, K., Alexakis, D.D., Sarris, A., 2013. Remote
Sensing for Archaeological Applications: Management, Documentation and Monitoring. In
Hadjimitsis, D.G., (Ed.), Remote Sensing of Environment: Integrated Approaches. InTech, pp.
57-95.
Hesse, R., 2010. LiDAR-derived Local Relief Models – a new tool for archaeological prospection.
Archaeological Prospection, 17, 67-72.
IGEAN (Innovative Geophysical Approaches for the Study of Early Agricultural Villages of Neolithic
Thessaly) Καινοτόμες Γεωφυσικές Προσεγγίσεις για την Μελέτη των Πρώιμων Αγροτικών
Εγκαταστάσεων της Νεολιθικής Θεσσαλίας. http://igean.ims.forth.gr/ (accessed 11.1.2015)
Intzesiloglou, Ch., 1995. Άγιοι Θεόδωροι. Θέση Αμπέλια. Αρχαιολογικό Δελτίο, 45(1990), Χρονικά
Β΄ 1, 204-205.
Intzesiloglou, Ch., 1997. Πρόδρομος-Μπουρντένια. Αρχαιολογικό Δελτίο, 47(1992 ΤΣΕΚΑΡΕ).
Χρονικά Β΄1, 235.
Intzesiloglou, Ch., 2000. «Έργα και Ημέραι» (1990-1997). Το Έργο των Εφορειών Αρχαιοτήτων
και Νεωτέρων Μνημείων του ΥΠ.ΠΟ. στη Θεσσαλία και την ευρύτερη περιοχή της (1990-1998),
pp. 373-379.
Intzesiloglou, Ch., 2010. Δυο αρχαίοι δρόμοι του Κιερίου. Σοφάδες, Β΄ Συνέδριο «Προσέγγιση στο
Παρελθόν του Τόπου», 75-79.
Krahtopoulou, A., Orengo, H.A., Palaiochoritis, K., Stamati, A. in preparation. Αναδασμός και
πολιτισμικό τοπίο στον Κάμπο της Καρδίτσας.

109

Menze, B.H., Ur J.A., 2012. Mapping patterns of long-term settlement in Northern Mesopotamia at
a large scale. Proceedings of the National Academy of Sciences, 19(14), E778-E787.
Milojčić, V., 1959. Ergebnisse der deutschen Ausgrabungen in Thessalien 1953-1958. Jahrhbuch
des Römisch-Germanischen Zentralmuseums, 6, 1-56.
Mylonas, G., 1928. Η Νεολιθική Εποχή εν Ελλάδι. Βιβλιοθήκη της εν Αθήναις Αρχαιολογικής
Εταιρείας, Athens.
Nikolaou, Ε., 2004. Αγρός Ευ. Κωστόπουλου. Αρχαιολογικό Δελτίο, 53(1998), Χρονικά Β΄2, 443.
Orengo, H.A., Palet, J.M., 2010. Methodological Insights into the Study of Centuriated Field
Systems: a Landscape Archaeology perspective. Agri Centuriati. International Journal of
Landscape Archaeology, 6(2009), 171-185.
Orengo, H.A., Ejarque, A., Albiach, R., 2014. Water management and land-use practices from the
Iron-Age to the Roman period in Eastern Iberia. Journal of Archaeological Science, 49, 265-275.
Palet, J.M., Orengo, H.A., 2011. The Roman centuriated landscape: conception, genesis and
development as inferred from the Ager Tarraconensis case. American Journal of Archaeology, 115
(3), 383-402.
Theocharis, D.R., 1967. Η Αυγή της Θεσσαλικής Προϊστορίας. Βόλος, Φιλάρχαιος Eταιρεία
Bόλου.
Theocharis, D.R., 1973. Νεολιθική Ελλάς. Αθήνα, Εθνική Τράπεζα Ελλάδος.
Tsountas, X., 1908. Αί προΐστορικαί Ακροπόλεις ∆ιµηνίου και Σέσκλου. Βιβλιοθήκη της εν
Αθήναις Αρχαιολογικής Εταιρείας, Athens.
Wace, A.J.B., Thompson, M.S., 1912. Prehistoric Thessaly. Being some Account of Recent
Excavations and Explorations in North-Eastern Greece from Lake Kopais to the Borders of
Macedonia. Cambridge University Press, Cambridge.
Wilkinson, T.J., 1993. Linear hollows in the Jazira, upper Mesopotamia. Antiquity, 67,
548-562.
Wilkinson, T.J., French, Ch., Ur, J.A., Semple, M., 2010. The Geoarchaeology of Route Systems
in Northern Syria. Geoarchaeology: An International Journal, 25(6), 745-771.