вход по аккаунту


The first thirty kilometres of the western front 19141918an aerial archaeological approach with historical remote sensing data.

код для вставкиСкачать
Archaeological Prospection
Archaeol. Prospect. 18, 57–66 (2011)
Published online 11 January 2011 in Wiley Online Library
( DOI: 10.1002/arp.397
The First Thirty Kilometres of theWestern Front
1914^1918: an Aerial Archaeological Approach
with Historical Remote Sensing Data
Department of Archaeology, Ghent University, Sint-Pietersnieuwstraat 35, 9000 Ghent, Belgium
During the first months of World War One, most armies realized the strength and possibilities of a new weapon: military
aviation carrying out aerial reconnaissance. Pilots and observers became the eyes of the army. Aerial photographs
were taken all over the different theatres of war, documenting a cultural landscape from which the relicts are often still
visible as scarsinthelandscape.These aerialphotographsare a massive and overlooked source for the archaeological
study of the Western Front in Europe.This paper describes how a specific, non-destructive approach and methodology
can provide new insights into extant and destroyed archaeological remains through combining GIS mapping and the
interpretation of thousands of historical aerial photographs. Copyright # 2011 John Wiley & Sons, Ltd.
Key words: Aerial photography; conflict archaeology; remote sensing for archaeology; First World War
After the first Battle of the Marne (5–11 September 1914),
the character of World War One changed from a moving
war to standstill in the trenches. Thus, the strength and
potential of a new weapon in the air were soon
recognized. Before long, pilots and observers became
the new eyes of the army, a role previously occupied
only by the cavalry and espionage (Carlier, 1921). Aerial
reconnaissance work was conducted by various nationalities and the resulting photographs survived in huge
quantities; approximately half a million of these aerial
photographs survived World War One and are kept
throughout the world (for the archival research on this
topic, see Stichelbaut and Bourgeois, 2009).
These extraordinary sources document a panEuropean landscape of horror that stretches from
the North Sea in Belgium to the French–Swiss border,
from the Black Sea in the south to the Baltic in the
north, comprising parts of Italy, the Balkans and even
more distant areas of the world. Aerial photographs
were taken all over these different theatres of war,
showing a cultural landscape from which the relicts
* Correspondence to: B. Stichelbaut, Department of Archaeology,
Ghent University, Sint-Pietersnieuwstraat 35, 9000 Ghent, Belgium.
Copyright # 2011 John Wiley & Sons, Ltd.
often remain visible as scars on the landscape, frequently
concealed from the untrained eye. These photographs
(see for instance Figure 1), used for the first time now as
primary sources, hold a wealth of useful data for
archaeologists, (war) historians and landscape specialists.
The intention of this paper is to further contribute to
the rapidly developing fields of twentieth century
conflict archaeology and research of conflict zones –
not only the areas in which the actual fighting took
place, but also areas towards the rear that had different
functions are included – as defined by Saunders (2009).
The archaeology of World War One, and twentieth
century conflict archaeology in general, is still in its
infancy but is a rapidly evolving field of research
(Saunders, 2002). Such an evolution is undoubtedly
being accelerated by the approaching centennial of
the start of World War One in 2014 and the massive
public interest in the subject. The current developments towards the protection and management of
World War One military heritage make it necessary to
investigate what this heritage originally comprises.
Specific research questions which require answers
are: What methodology should be followed to enable
a desktop assessment of the conflict landscape? How
can any possible material remains be located in detail?
What is their diversity and distribution? Is it possible
to analyse an entire landscape by means of aerial
Received 23 October 2010
Accepted 3 December 2010
B. Stichelbaut
of field defences from the North Sea to Switzerland
begins at the beaches of Nieuwpoort and Lombardzijde, but perhaps most important because this is the
first area that has been systematically researched using
a large number of historical aerial photographs.
Historical background
Figure 1. Belgianaerialphotograph (17 March1917) showingtheruined
city of Dikmuide, German trenches and watermarks of the cities postmedievalfortifications.(Source:KLM-MRA ¼Koninklijk Legermuseum
- Muse¤e Royal de l’Arme¤e¤, Brussels, Belgium.)
photographs or can than this be used only for small
case studies (Stichelbaut, 2006)? What kind of information would this provide?
As regards these issues, a multidisciplinary aerial
photographic study can provide more adequate answers
than any other source. The long-distant views provided
by aerial photographs enables a redirection of the focus
of World War One archaeology from a site-directed
approach to research on a landscape scale.
Study area
A large study area of approximately 445 km2 was
selected in order to examine the (conflict) archaeological application of World War One aerial photographs. This area is roughly situated at the Western
Front between the North Sea coast at Nieuwpoort and
Houthulst in the south (Belgium). The research area
can be defined as the first 30 kilometres of the Western
Front. This is not only because this continuous line
Copyright # 2011 John Wiley & Sons, Ltd.
It is not the intention to provide a full and detailed
account of all the battles and minor actions that
occurred in the study area. Yet, it is necessary to
provide some background information on what
happened during World War One in the area and
how the front line came into being. On 4 August 1914,
German troops crossed the Belgian border as part
of the Schlieffen Plan to overrun France with an
outflanking movement through neutral Belgium. The
forts around Liège and Namur did not stop the
invaders and the Belgian army retreated to reduit
national Antwerpen (Antwerp). It soon became clear
that the fortresses around the city were unable to
provide adequate protection so the army drew further
back while Antwerp surrendered on 9 October 1914.
The Belgian Army reorganized behind the IJzer River,
which was the only natural line of defence in the
province of West Flanders. On 18 October, the Battle
of the IJzer River began (18–30 October 1914) and
options were explored to initiate strategic flooding
of the river to halt the Germans. From the end of
the Battle of the IJzer onwards, the sector remained
quiet compared to other parts of the Western Front.
The major offensives in the following years, such as the
Second and Third Battle of Ypres, all took place to
the south near the Ypres Salient. Because this front
stayed largely immobile, it is an ideal testing ground
to investigate how static trench warfare evolved from
the early beginnings in 1914 and 1915 to the final
phases of the conflict in September 1918.
Data acquisition
Archival research at different institutes resulted in
a detailed overview of what aerial photographs
could be found in the archives. A total of 6,591 of
these photographs were selected and these contact
prints were scanned at a resolution of 400 dpi. The
photographs date from all periods of the war. Early
aerial photographs were selected where possible but
these are in the minority because there are considerably less photographs available for the earlier stages of
the war. The earliest aerial photograph found in the
archives dates from 30 April 1915.
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
The First Thirty Kilometres of the Western Front 1914–1918
The amount of information on the selected photographs could only be managed, stored and analysed in a
database. Fundamental data about the photographs, such
as date, flying height, nationality and, most importantly,
the original unique photograph number were recorded
to permit analysis of the data collected. The majority of
the selected photographs were taken at an altitude of
between 1,250 and 2,750 m and are near vertical.
The aerial photographs were georectified using a
second order polynomial transformation. Between ten
and 15 ground control points (GPCs) were indicated.
This often caused problems because of the differences
in landscape depicted on the aerial photographs and
on the more recent cadastral maps which were used as
main reference data. Most of the GCPs are field
boundaries that were visible on both sources.
In total 6,591 photographs have up to now been
georectified and made accessible through a GIS file
outlining the location of the aerial photographs (see
Figure 2). The located aerial photographs form a
continuous study area with a high density multitemporal coverage of aerial photographs.
provide a record of older sites that have since then
disappeared. The images proved to be an excellent
source for the detection of numerous watermarks,
most of them being medieval moated sites (Bourgeois,
2003; Bourgeois and Stichelbaut, in press). The
photographs are therefore a privileged source for
the study of these sites because circumstances during
the war provided the ideal (humid) conditions for the
detection of many sites.
Second, the photographs date from before the period
of large landscape changes that took place after the
Second World War (Antrop, 2007). Therefore, it was
possible to record information about many archaeological sites that have now already been destroyed,
levelled to the ground or built on.
Third, the images are important as a photographic
record of the landscape at the beginning of the
twentieth century. For the first time in history there
is an almost continuous coverage of aerial photographs
for a large part of Europe.
The conflict landscape of World War One
Aerial photograph interpretation –
traditional archaeology
Although the photographs are taken in a context of
military intelligence during World War One, they also
The archival research to obtain a continuous coverage
of aerial photographs was successful and the mapping
of all visible features is based on the georectification
of these primary sources. The detailed study and
interpretation of aerial photographs resulted in a GIS
Figure 2. Density of 6591georectified aerial photographs (each square corresponds with a single historical aerial photograph).
Copyright # 2011 John Wiley & Sons, Ltd.
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
layer of almost unique 27,000 features, the majority of
which relate to World War One.
The area studied is literally and symbolically the
first 30 km of the Western Front. In many respects, this
is a unique research area and consists of a newly
constructed military landscape that came into being
after the Battle of the IJzer and was occupied by both
German and Allied forces until late September 1918.
Therefore, the ideal circumstances were created to gain
insight into a complex conflict landscape from the
Allied first line to no-man’s-land up to far behind
the German lines. Second, this is important because it
comprises different landscapes: from coastal dunes, to
a flat and polder landscape, and finally a more closed
landscape which undulates in the east and south of
the study area. This area was defended by Belgian
divisions on one side with, temporarily, British and
French divisions in the north and south and, on the
opposite side of no-man’s-land, German troops.
Aerial photographic interpretation and subsequent
GIS mapping procedures resulted in a GIS layer
containing a detailed account of more than 26,500
features of World War One (see for instance Figure 3).
It was surprising to see a large typological diversity
of many feature categories, such as artillery emplacements and fire and communication trenches. This
diversity is much wider than at first was expected
by examining contemporary field manuals for the
construction of trenches. German manuals, in particular, are poorly endowed with descriptions and plans
of trenches (Prussian War Ministry, 1916; Chef
des Generalstabes des Feldheeres, 1918). Comparison
between both sources revealed that there is a clear
distinction between what is visible on aerial photographs (and thus what really happened on the ground)
and how trenches should be built according to
field manuals (representing an idealized battlefield
Small-scale application
There are two different applications of the aerial
photographic approach. The first application is based
on small research areas where the research goals are
dominated by site specific research questions: What
happened on this particular site? What features are
potentially preserved beneath the surface? What are
the distribution, density and importance of features?
What kind of narrative can be formulated with this
The interpretation of a sequence of time-based aerial
photographs and the mapping of all visible features
Copyright # 2011 John Wiley & Sons, Ltd.
B. Stichelbaut
resulted in a detailed GIS inventory with temporal
data (terminus post quem, terminus ante quem and date of
destruction of war features), which is extremely useful
for the study of particular sites. Due to the high
temporal resolution that is obtained –some areas
were photographed on up to 30 different dates – it is
possible to date individual features that are found
during archaeological fieldwork. This strategy evidently supports the work of local heritage councils. For
small case studies there are many opportunities for
interdisciplinary research, especially combining aerial
photograph interpretation and an array of geophysical
prospection techniques. Along the former Western
Front the first experiments to detect World War
One features took place using resistance survey,
magnetometry (Bossuet et al., 2001; Masters and
Stichelbaut, 2009) and electromagnetic induction
(Stichelbaut et al., in press). These methods provide
results to evaluate where material remains of the
conflict are still preserved beneath the surface. The
sheer scale of the Western Front does not allow a total
geophysical evaluation, whereas aerial photograph
research is possible for large research areas of several
hundreds of square kilometres (see below).
Figure 4 shows two aerial photos of the same
location, photographed on two different dates. The left
image was taken on 30 September 1916, whereas the
right-hand image was photographed more than one
and a half years later on 2 April 1918. This complex lies
near Langemark and up to the start of the Third Battle
of Ypres it was located almost 4 km behind the first
lines. Consequently, in this period many gun positions
were spread across the immediate vicinity because the
site was located in an artillery zone. In Figure 4A we
can observe a German exercise trench system with
artificial shell holes. These fake shell holes were
constructed to make the practice area more realistic
and provide the training infantrymen with an artificial
No Man’s Land. An exact date for the construction of
these trenches is not available, but 30 September 1916
can be considered as terminus ante quem. The distance
from the front line provided some security but it was
definitely not out of reach of the Allied artillery. This
might have been intentional, to enable newly arrived
soldiers to become accustomed to the continuous and
awful battlefield noise.
Following the start of the Third Battle of Ypres, the
front line edged closer and closer until, at the end
of October 1917, the area became occupied by Belgian
troops (B). By this time, the landscape had been totally
transformed and many of the early features had
become completely invisible. Only the most pronounced features, such as the breast-work trench
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
The First Thirty Kilometres of the Western Front 1914–1918
Figure 3. Exampleofthedetailedlevelofthe GISmapping, representingfourdifferentphasesintheevolutionofthe Germanfrontlinenear Diksmuide.
and parts of the road network, are recognizable. The
function of this place on the battlefield changed
significantly. From a relatively safe place behind the
lines where German troops practiced infantry tactics,
it became the front-line area. The former practice
Copyright # 2011 John Wiley & Sons, Ltd.
trenches were incorporated into the theatre for
real fighting and were occupied by Belgian troops.
A number of Belgian advanced posts (annotated as
P.P. on the photograph) can be observed in
Figure 4B. The new front line, indicated by a white
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
B. Stichelbaut
Figure 4. Comparative study of photographs taken on different dates. (Source: aerial photographs,KLM-MRA¼Koninklijk Legermuseum - Muse¤e
Royal de l’Arme¤e¤, Brussels, Belgium.)
dotted line (on the original aerial photograph), was
deliberately abandoned by the Allies in April 1918
because of the German Spring Offensive. The area
subsequently became part of the German hinterland
and changed its function once again for a couple
of months until the Final Offensive in September 1918.
This example clearly shows the need for a comparative
study of aerial photographs to acquire a full understanding of the context and meaning of a small area
on the battlefield.
Large-scale application
The broad view of the aerial photographs and the large
quantities available allows a reorientation towards a
diachronic landscape scale of research. For the first
time in history thousands of aerial photographs are
combined to gain insight into the structure of the
conflict landscape, not only for the combat areas at
the front, but also far behind the lines to see how the
military occupation made use of this landscape.
The detailed inventory of polygons was converted
to point and line features in order to enable the
landscape scale research. A selection of some important feature categories is represented on Figure 5.
This overview of war features gives an idea of the
distribution and density of the features recorded from
the aerial photographs. This approach enables one
Copyright # 2011 John Wiley & Sons, Ltd.
to analyse how the conflict landscape was really
organized. The map illustrates the distinct locations of
different categories of features, such as the occurrence
of trenches at the frontline, zones that were only
used for artillery emplacements, areas with massive
amounts of barracks, training features in the rear, etc.
Table 1 represents the industrialized scale of the
conflict and represents the total amount of mapped
features in the area studied.
In the case study of the first thirty kilometres the
distribution of German military features in the German
occupied part of the frontline was analysed, leading to
some interesting insights into the spatial organization
of the landscape. Figure 6 shows where the largest
concentrations of German features are to be found
in relation to the German/Allied front during the
period between the Battle of the IJzer and July 1917.
This figure clearly indicates that a number of feature
categories are located in specific places on the landscape and that their distribution is distinct. The
frequency distribution graph shows that the majority
of the trenches are located in the area between the front
and 5 km to the rear. A large number of bunkers
are also located at the front, and barbed-wire
entanglements are recorded most frequently at the
front. Yet there are large peaks towards the rear,
corresponding to the German second and third
positions. The majority of the artillery emplacements
are located between 2 km and 4 km from the German
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
The First Thirty Kilometres of the Western Front 1914–1918
Figure 5. Overview map of most important categories of mapped features. This figure is available in colour online at
front. This artillery zone is adjoined by an area that
consists mainly of barracks. The graph shows a large
peak near 4.5 km but from then on, this line gradually
decreases. The distribution graph shows that at a
distance of 13 km from the front almost no features
have been recorded.
In the area studied, almost 670 km of trenches and
4.6 km of saps (approach trenches oriented towards the
enemy lines) were recorded. In addition, there are
Copyright # 2011 John Wiley & Sons, Ltd.
1,718 bunkers, 2,677 artillery emplacements, 3,292
barracks, and many more sites. This impressive
number points to the massive impact of military
activity during World War One. The GIS mapping of
war features enables us to examine what types of
trenches were favoured and which were rarely used. It
is impossible – and not advisable from a cultural
heritage management point of view – to protect all of
them or fully excavate every one. In the near future, it
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
B. Stichelbaut
Table 1. Total length and number of mapped sites in the area studied.
will be necessary to consider carefully constructed
desiderata for the protection of this specific type of
heritage. The digital maps that have been developed
are, without doubt, a valuable starting point.
Obviously, this does not mean that all these features
have an archaeological footprint in the soil. This is
precisely one of the advantages of using historical
aerial photography. The careful investigation of the
aerial photographs enables the mapping of what was
present in the landscape during World War One.
Features such as canvas screens to conceal ground
movements, wire obstacles and, to some extent, even
barracks will probably not be preserved in the soil
because they barely made an impact. The aerial
photographs and the GIS mapping show what was
constructed during the war rather than what is still
preserved beneath the surface. Many inventories
start from what is still preserved in the present-day
landscape (Decoodt, 2007). These are certainly very
useful documents but focus on the features that are
still visible in the present-day landscape – the aerial
photographs and their analysis allow research into all
aspects of the historical landscape.
Copyright # 2011 John Wiley & Sons, Ltd.
The detailed analysis of the feature categories
illustrates the individual character of smaller regions
within the study area and the strong relationship
between the network of military features and the
landscape. For instance, the absence of a large trench
density near the inundation, the location of defensive
positions west of the villages, the construction of
trenches with breast-work trenches in the polders and
wet areas, the importance of the Wijnendale plateau
and the Klerken-Staden ridge for the construction
of the German positions, the appeal of forests in which
to construct barracks, and the presence of a large
number of canvas screens on the polders can be quoted
in this respect. Positions were built for their strategic
advantages and the protection provided by the
landscape (Strubbe, 2006). It is important to note
that the location of lines of trenches and defensive
positions is not always as might be expected when
studying German war regulations. In some cases, the
aerial photographic interpretation shows that the real
situation on the battlefield is sometimes different from
the ‘ideal battlefield’ referred to in the German war
manuals. For example, the location of the German
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
The First Thirty Kilometres of the Western Front 1914–1918
Figure 6. Distance of German features from the German frontline.This figure is available in colour online at
second and third positions in front of the StadenHouthulst ridge and the Wijnendale plateau contradicts what is proscribed to by the German military
doctrine. The positions are built into the slope facing
the enemy and prone to ground observation – instead
of on the backward slope east of the ridges.
A large number of aerial photographs have been
selected from the archives to be able to apply the
methodology developed to research a large area in
West Flanders (Belgium) between Nieuwpoort and
Houthulst. This proved to be an ideal case study for
establishing what information the aerial photographic
interpretation and mapping of visible features can
In particular, the use of aerial photographs proved to
be very important for the rapidly developing field
of conflict archaeology. All the more so because
this quickly developing and relatively new discipline
shows an evolution from the site-centred approach of
amateur archaeologists mainly interested in recovering interesting finds (often referred to as ‘battlefield
archaeology’, see, for instance, Laffin, 1987), towards
a more holistic multidisciplinary research perspective
concerning the study of material remains and landscape of World War One. The physical remains of
conflict have long been ignored in professional and
scientific archaeology (Sutherland and Holst, 2005).
By means of an aerial photographic approach, a
contribution can be made to gain further insight into
the impact of war on the landscape and the archaeological records of this violent episode in European
Copyright # 2011 John Wiley & Sons, Ltd.
history. The aerial photographic research of conflict
landscapes (not only World War One, but by extension
all twentieth century conflict archaeology) should be
incorporated as an integral part of an interdisciplinary
approach, intrinsically bound to the anthropological
archaeology of conflict zones.
The aerial archaeological research of the Western
Front using contemporary aerial photographs provides an accurate insight into the density, distribution
and diversity of war features. This approach enables
an analysis of how the conflict landscape was
organized and where certain types of features were
situated, providing a detailed level of information
that cannot be found in any other historical source.
By means of point density maps and spatial analysis
methods, several elements have been combined in
order to gain a better insight into the organization and
structure of this conflict landscape.
Antrop M. 2007. Perspectieven op het Landschap. Academia
Press: Gent.
Bossuet J, Camerlynck C, Brehonnet C, Petit C. 2001.
Magnetic prospecting of diachronic structures (antiquity to First World War) on the site of the sanctuary
of Ribemont-sur-Ancre (Somme, France). Archaeological
Prospection 8: 67–77.
Bourgeois J. 2003. Les Sites Fossoyés Médiévaux de la
Région de Comines-Warneton (Province de Hainaut,
Belgique). Revue du Nord 85(353): 141–159.
Bourgeois J, Stichelbaut B. In press The use of historical
aerial photographs in West-Flanders for the detection
of archaeological sites. In The Spy in the Sky: Archaeology
from Aerial Archives, Hanson W, Oltean I (eds).
Cambridge University Press: Cambridge.
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
Carlier A. 1921. La Photographie Aérienne Pendant la Guerre.
Librairie Delagrave: Paris.
Chef des Generalstabes des Feldheeres. 1918. Allgemeines
über Stellungsbau. Druckerei des Chef des Generalstabes
des Feldheeres; S.l.
Decoodt H. 2007. De Sporen van ‘‘Den Grooten Oorlog.
Monument, Landschappen en Archeologie 26(1): 4–36.
Laffin J. 1987. Battlefield Archaeology. Ian Allen: Hersham.
Masters P, Stichelbaut B. 2009. From the air to beneath the
soil – revealing and mapping Great War trenches at
Ploegsteert (Comines-Warneton), Belgium. Archaeological Prospection 16: 279–285.
Prussian War Ministry. 1916. Manual of Position Warfare
for All Arms. Part I: The Construction of Field positions
(Stellungsbau). Translated from German, and issued by
the General Staff. S.n; Berlin.
Saunders N. 2002. Excavating memories: archaeology and
the Great War, 1914–2001. Antiquity 76: 101–108.
Saunders N. 2009. Conflict Archaeology. http://www. (Retrieved
30 January 2009).
Copyright # 2011 John Wiley & Sons, Ltd.
B. Stichelbaut
Stichelbaut B. 2006. The application of First World War
aerial photography to archaeology: the Belgian images.
Antiquity 80(307): 161–172.
Stichelbaut B, Bourgeois J. 2009. The overlooked aerial
imagery of World War One: a unique source for
conflict and landscape archaeology. Photogrammetrie Fernerkundung – Geoinformation 3: 231–240.
Stichelbaut B, Saey T, Meeuws F, Bourgeois J, Van Meirvenne M. In press. World War One Heritage in Belgium:
combining historical aerial photography and EMI. In:
Cowley D. (ed.). EAC Occasional Paper No. 5/
Occasional Publication of the Aerial Archaeology
Research Group No. 3. Archaeolingua: Budapest.
Strubbe L. 2006. Een Lappendeken met Groene Naden:
het Landschap van de Ieperboog ‘Anno 1914’. In De
Laatste Getuige. Het Landschap van Wereldoorlog 1 in
Vlaanderen, Chielens P, Dendooven D, Decoodt H
(eds). Lannoo: Tielt; 147–152.
Sutherland T, Holst M. 2005. Battlefield Archaeology. A
Guide to the Archaeology of Conflict. English Heritage:
Archaeol. Prospect. 18, 57–66 (2011)
DOI: 10.1002/arp
Без категории
Размер файла
801 Кб
front, data, approach, archaeological, thirty, aerial, remote, kilometres, first, historical, sensing, western, 19141918an
Пожаловаться на содержимое документа