Kemmelberg

For the sixth and fifth century BC period, there are currently about twenty fortified hillforts which have been identified in continental Europe which were inhabited by local elites.

The Kemmelberg occupies an eccentric position, a little remote to the north-west, towards the British Isles, where thousands of hill forts are distinguished. Some of these differ in character from the continental altitude-sites of this period.

Central fortified altitude-sites, 6th-5th centuries BC (black dots are continental European sites, blue triangles are related Mediterranean places with Greek or Etruscan roots).

During the Iron Age the Kemmelberg was provided with a complex system of ditches, ramparts, and wooden palisades.

Mainly due to the destruction of the First World War, but also because of the difficult legibility of the soil, only a very fragmentary picture of the fortifications has so far been discovered.

1 From LASER and LIDAR to DTM and LRM

In 2008, the Kemmelberg was scanned by Eurosense from an aircraft which was equipped with LIDAR systems ('Laser Imaging Detection And Ranging').

This resulted in an accurate 'Digital Terrain Model' (DTM). All objects on the earth's surface, such as trees and buildings, were digitally removed, so that only the bare terrain remained.

LIDAR determines the distance to a surface using laser pulses. The laser scans the surface. The height of points on the surface can be determined from the information taken from the reflected pulses. When provided with the positioning and correction for the movements of the aircraft, it is possible to arrive at a precise digital topography.

There are three operations for a LIDAR scan recording: LRF ('Laser Range Finder') to measure distance with a laser, GPS to determine the position of the aircraft, and INS ('Inertial Navigation System') to measure aircraft movements.

The principles of LIDAR.

The DTM can then be coloured digitally for each height segment, and the sun's position can also be artificially adjusted so that shadows can help to visualise any irregularities. Highest elevation levels are coloured bright red, lower elevation levels phase through dark red over brown, and yellow to green. The level difference in the photo between top edge (bright red) and bottom edge (light green) is about 100m.

3D LIDAR DTM image, Kemmelberg.

In the DTM image, the plateau area is coloured bright red. On the northern side of the plateau, the connecting steep edges in a dark red shade are nicely visible. Between the two white arrows is the only Iron Age ditch which can be seen on DTM images.

The steep edges enclose an area of ten hectares, about the maximum outer limit of the Iron Age fortress. The plateau area is slightly uneven and approximately 4.5 hectares in size. The most important residences were concentrated within this area. The plateau consists of an east-north-east-orientated longer part (of three hectares), with traces of compartmentalisation.

The smaller west-south-west part (1.5ha) is rather oval in shape and slopes down to the west. Vague indications of post holes for rectangular houses have been found in that part.

With LRM ('Local Relief Models') it is possible to represent local, small-scale height differences thanks to LIDAR results. The colour code being used is as follows: yellow is higher than the immediate vicinity, blue is lower than the immediate vicinity, red is in-between. In this way one can better 'see' the divisions, both of the plateau and in the plateau, and also recognise the steep edges. As mentioned, the only Iron Age ditch so far detected is also visible.

The purple outline is an interpretation of the maximum area of the fortifications, including steep slopes and waste layers. The green interpretation line follows the inhabited plateau, consisting of two parts, and showing a vague compartmentalisation in the eastern part.

This was done by combining LRM with the presence of centennial trees (trees over a hundred years old). This made it possible to locally reconstruct the relief from before the First World War. Apparently the compartmentalisation is perpendicular to the outer edge, with intervals of around 35m. It is not certain that this compartmentalisation dates from the Iron Age: it could be younger or even older...

LIDAR LRM image, Kemmelberg. The distance between the white stars is 450m.

The twenty-or-so altitude-sites on the continent from this period, are difficult to compare. Mont Lassois in France is 350km to the south-east of the Kemmelberg as the crow flies. It is also a witness hill, whose inhabited plateau area (Mont Saint-Marcel) of five hectares (400m long by a maximum of 150m wide) is comparable in size to that of the Kemmelberg.

The plateau is clearly compartmentalised. The extensive ramparts and ditch complexes (black thick lines and brown-red lines) are striking to see. They extend far beyond the plateau part.

The River Seine runs 100m lower at the foot of the hill. The finds are overwhelming, while also being a few generations older than those of the Kemmelberg. Striking are the ramparts and ditches which are up to 400m from the plateau part. This fact shows that the wide Kemmelberg area can still reveal its surprises.

Map of the Mont Lassois area.

2 Northern rim plateau

The majority of research could be carried out on the northern edge of the Kemmelberg plateau. The north-east to north-west edges are very steep. Mass excavations have been identified in the north-to-north-east segment of the fringe zone. Large chunks of iron sandstone were also found in the archaeological layers, possible evidence of mining for iron production and/or for use in the construction of ramparts.

Large piece from an ironstone bank, in the middle of an archaeological layer.

Apparently, ramparts, ditches, and palisades were also moved and updated several times. This happened at the top of the steep slope, with canals which were sometimes ten metres wide and between four to five metres deep. An access road could also be identified on a certain strip.

Traces of fortification and access road: 1) wall; 2) ditch; 3) excavated zones with possible palisade or terrace extension; and 4) access road.

Wooden palisades were dug into clay packs, perhaps for better preservation and positioning. Dozens of cubic metres of clay packs were moved up from at least ten metres below to prop up the palisade posts, perhaps also as a clay reserve for pot-baking.

Conversely, there was a waste dump over a distance of at least fifty metres, under the northern rampart of at least a hundred cubic metres, and this after composting (Photo 2a).

To the far right, on top of the rampart (the kink between slope and plateau - see Photo 2b), there is an almost-horizontal zone, characterised by the presence of several filled-in ditches and the remains of palisades, especially on the side of the central plateau (Photo 2c).

Photo 2a: waste dump at the left, at the foot of the northern steep slope.

Photo 2b: on the rampart.

Photo 2c: on the top of the rampart is an almost-horizontal zone (plateau).

3 The south-west end of the plateau

At the south-west end of the plateau was a simple closing-off ditch with earthen wall and palisade.

The kink in the landscape, where the ditch used to be, is still visible in the landscape. At the top, the palisade and earthen wall have completely disappeared. The earthen wall has eroded away and is now almost completely filling the ditch.

Closing-off ditch with earthen wall.

On the 2D representation of the electrical resistance measurement, the zone with the closing ditch is nicely delineated (in the light blue band, in the centre of the image). The measurement shows the opposite direction to that of the photo.

Electrical resistance measurement - 2D view (cross-section) of the terminating ditch area.

4 Water and waterway

Water is, and was back then, also vitally important. No water wells were found during the excavations, which can be explained by the presence of sources on the hilltop area. A source level is situated around the 140m height line, the contact zone between the sandy ironstone top layer and the thin clay layer underneath. First World War damage may have broken through the impenetrable clay layer here and there, which led to less pronounced source flows.

Some of these small sources still supply discontinuous small water flows. They are located on the southern edge of the plateau, at the head of a small asymmetric valley in the forest which evolves towards the Douve brook, two kilometres further on and a hundred metres lower. The Douve brook used to be navigable for small barges. It runs ten kilometres further downstream into the River Lys at Warneton (Waasten in Belgium).

The Lys continues to Ghent where it flows into the Scheldt. The Scheldt flows via Antwerp into the North Sea towards the south-western end of the Netherlands. This was clearly an important shipping route in the Iron Age. Presumably, the 'road' from the sources at the top to the Douve played a strategically important role.

The current sailing distance from the Douve to Antwerp is 170km via the Douve itself, the Lys, and the Scheldt. With the many extra meanders which would have existed prior to more recent remedial works and alterations, this must have amounted to more than 200km in total distance in the Iron Age.

Small source at the top, with concrete cover.

Like other sites, the Kemmelberg also appears to consist of a complex system of ramparts, ditches, and palisades. How far they extend beyond the plateau is unknown. The whole must have had the appearance of a (fortified) 'ranch' rather than a 'castle'.

There may have been compartments for some notable housing units of the ruling elite which lay close to the northern edge of the plateau. The exact location of any warehouses, workshops (pottery, metalworking, and so on), livestock and horse fencing, etc, is anyone's guess.

Water sources on the southern edge of the plateau provided the water supply. Clay, which was required for pottery manufacture, was found somewhat lower. Large stocks of this clay were found at the top of the plateau's edge.

To date, no traces of violence, weapons, or destruction by fire have been found.

 

REFERENCES:
Lock, Gary & Ralston, Ian (Eds), Hillforts: Britain, Ireland and the Nearer Continent, 2017

Fernández-Götz, Manuel, Urbanization in Iron Age Europe: Trajectories, Patterns, and Social Dynamics, J Archaeol Res (2018) 26:117-162

Van Doorselaer, A, Putman, R, Van der Gucht, K, & Janssens, F, De Kemmelberg een Keltische bergvesting, Westvlaamse Archaeologica, Monografieën III, Kortrijk, 1987 (in Dutch) (see external link, below)

RELATED KEMMELBERG LINKS:
Geology: Witness Hills
Landscape
First World War: Relics & Monuments
First Elite
Photo Focus: First Elite

EXTERNAL LINKS:
De Kemmelberg: een Keltische bergvestiging (in Dutch)
Inventory of 4,147 hillforts in Britain and Ireland
Eurosense, Geo Information Services

RELATED HISTORY FILES LINKS:
Prehistoric Britain
Early Cultures of Europe
Celts
Modern Belgium
Modern Netherlands

 

Text copyright © Archeo Kemmelberg. An original feature for the History Files: Kemmelberg.