Ancient Roads and GPS Survey: Modelling the Amarna Plain
Fenwick, Helen, Antiquity
The construction of the city now known as Tell el-Amarna was begun in the fifth year of the reign of the Pharaoh Akhenaten (c. 1350 BC) (Kemp 1989). The city remained the centre of power during Akhenaten's reign, but upon his death power returned to the older centres. A sparse population probably remained in the city during the reign of his successor, Tutankhamen, but after this, the city was abandoned, although a couple of small settlements remained (Kemp 1989). The city of Amarna therefore allows the archaeological investigation of a wide area without the encumbrance of subsequent occupation.
The city is located on the Nile, about halfway between the important centres of Memphis and Thebes. It lies mainly on the east bank, on a flat piece of desert which is surrounded by a crescent-shaped escarpment. In the central part (Figure 1) is a temple and ceremonial centre, with surrounding accommodation for the city's officials, workers and citizens. The escarpment itself contains rock-cut tombs attributed to the more prominent inhabitants. Akhenaten marked the limits of his city by placing boundary stelae both on this eastern escarpment, and on the distant escarpment on the western bank of the Nile. In the area between the main city and the eastern escarpment lies an area of desert which at first viewing would appear to be devoid of activity. It is on this area that the new survey work reported here was focussed.
[FIGURE 1 OMITTED]
Flinders Petrie conducted the first archaeological investigations at Tell el-Amarna between 1891 and 1892 (Petrie 1894), although the site had been previously surveyed and included in the Napoleonic Description de l'Egypte (Kemp & Garfi 1993). Petrie noted that the site had been recently looted and remarked with a slight irony that his work was "better late than never" (1894: 2). Within the seemingly blank area between the central city and escarpment, Petrie recorded a series of roads (1894: 4). These 'roads' were simply strips where the stones that litter the surface had been cleared to either side to produce an accessible routeway through the desert. Petrie began to record them using a compass and fixed points on the escarpment, and by "counting steps all day" (Petrie 1984: 1). Unfortunately an injury to his foot cut short this exercise, so it was left to his assistant, Howard Carter, to complete the record. On his return to England, Carter then posted his notes and maps of the road system to Petrie, but they never arrived (Petrie 1894: 4). Later investigators at the site also attempted to record the road system; for example Davies (1906), whilst recording the rock-cut tombs, and Timme (1917), a military surveyor with the German army, who produced an impressive map of very straight roads criss-crossing the Amarna plain.
The Egypt Exploration Society originally undertook work at the site between 1921 and 1936, and the current programme of research began in 1977 (Kemp 1989). The area of the central city has been surveyed in detail and published (Kemp & Garfi 1993). With the advent of high accuracy, differential Global Positioning Systems (GPS) the total mapping of the large area contained by the eastern escarpment has now become a practical proposition.
The Eastern area GPS survey 2001-3
A small area of the escarpment was surveyed using an EDM in 1996; including the area of the North Tombs, and it was felt that this was a good point to start the new survey in 2001 which has made use of a GPS. The two main aims of the survey are to build a topographic model and to locate any known or previously unrecorded sites in the Amarna hinterland. In the first season the survey was undertaken using a Geotronics Geotracer System 2000 L1-RTK differential Global Positioning System (GPS). Subsequently, the survey has been undertaken using a Leica 500 series differential GPS. The GPS equipment receives and compares radio signals from 24 satellites which in this case are under the control of the U.S. Department of Defense (Leick 1995). Various factors can affect the accuracy of the system, including selective availability, multi-path and ionospheric and atmospheric interference (Heywood et al. 2002). Until May 2000, a deliberate error was introduced by the American military for security reasons which meant that any single reading from a receiver had an effective accuracy of only 100 m for 95 per cent of the time. With the removal of this 'selective ability', the accuracy is now increased to 25 m for 95 per cent of the time (Heywood et al. 2002).
The accuracy can be further improved by using two receivers, one static and one roving, the set-up known as differential GPS. The roving receiver is placed upon the point which needs to be recorded and a reading taken, and the readings are calibrated with those of the static receiver (which needless to say should not vary). Although slightly more time-consuming than taking automatic interval readings, the accuracy achieved is far greater (c.f. Branting & Summers 2002).
The advantages of GPS survey over other methods have been discussed elsewhere (Chapman & Van de Noort 2001; Branting & Summers 2002). In addition to the speed at which a large area can be surveyed, it is a great advantage in broken country to be able to position a point without line of light. The main disadvantage is the need to maintain a clear view to the sky above the receiver so that it can communicate with the satellites. This means that survey cannot be undertaken close to buildings or under trees. At Amarna, this inhibited survey in those areas that are close to the escarpment, and in the vicinity of the rock-cut tombs.
In total over 80 000 points have been recorded in three seasons at Amarna. Point data have not been taken on any formal grid system, but instead have been recorded at paced intervals for certain features and areas, or along specific topographical features, such as the bottom, middle and top of slopes. This method of survey has been developed in response to several surveying projects undertaken in England, on a variety of sites, and has been shown to produce excellent results (Chapman & Fenwick 2002; Chapman & Van de Noort 2001; Fenwick 1999, 2000). A network of fixed points were established at various locations across the Amarna plain and placed on immovable objects so that the curiosity does not lead to their removal. The network of fixed points was re-surveyed in 2003 in order to provide a control against the introduction of any systematic errors between the three seasons of survey. The total area surveyed so far covers 15 km2, with the maximum north-south length being 7.5 km and maximum width being 3 km (Figure 2). Data points have been logged and processed using Leica SkiPro software. Once processed the data have then been imported into ArcView GIS software. Within this software package modelling of the topographic surface has been undertaking using 3D Analyst.
[FIGURE 2 OMITTED]
The main results of the research to date has been to produce a detailed topographical model of the area and to map the ancient road network seen by Petrie and partially recorded by Davies (1906) and Timme (1917) (Figure 2, 3). The roads, consisting of areas of cleared stone, are of differing lengths, widths and varying degrees of visibility. Variations in width have been recorded in the range of 2-11 m. In some locations the edges of the roads were lined with larger stones, in other areas the edges are simply demarcated by the difference between densities of small stones (Figure 3). They are rarely visible from the air, and many are invisible on the ground when viewed in the wrong light. In some areas they have been badly damaged by later wadi activity or more recent vehicle track marks. Therefore, the survey has shown that the best method of recording is to plot the edges on the ground, in the best light conditions, for different sections of road network. Points were taken along the edges of the roads visible on the surface, at paced intervals equivalent to c. 5 m. Breaks in the roads were noted, whether this is due to an actual break in the road system or to the fact that a section has been removed or obliterated in more recent times. Where roads meet, their relationship has been recorded, whether they met at a junction or whether one road overlies another, and is therefore later. In total 70 separate roads have been recorded with a total length of 30 km.
[FIGURE 3 OMITTED]
Initial study of the roads suggests that they served differing functions. This observation was first postulated by Petrie who divided the roads into categories, such as patrol roads, and those which ran to quarries, tombs and the stelae. The patrol roads are described by Petrie as crossing the plain, running along the escarpment and into the desert. Due to the nature of the slopes that they transverse, Petrie concluded that they must have been for foot patrols, as many would have been too steep for chariots. This has been confirmed by the current survey which has shown that in some areas it would appear that it was more important to continue along a set route, than to take a more logical course that respected the dominant topography. Two roads have been recorded taking a straight, unwavering course directly up the escarpment. There are many other, more circuitous routes, by which to transverse this escarpment but the roads avoid these and take a direct, straight route to the top. Elsewhere, roads avoid flat desert and traverse small outcrops that appear along their route. The idea of foot patrols using these roads is also recorded in one of the rock-cut tombs. The tomb of Mahu, chief of the Amarna police, has depictions of men running along roads (Davies 1906).
This is not to say that all the roads were completely straight. A number have turns and kinks, and the main exceptions are the roads encircling the sites known as Workmen's Village and Stone Village (Figure 2). Here the road systems seem to run around the edge of the settlements, and though their precise function is unknown, they could possibly have acted as some form of control mechanism to keep the residents in or to keep others out (Kemp 1989). Once the survey of the roads is completed, functional aspects will be explored in greater detail. As well as the patrol roads, there is a network of roads running directly from the central city to the north and south tombs. One line of enquiry will be to see if is it possible to link the roads from specific tombs to specific houses in the central city, and hence suggest family links between residents of the city and occupants of the tombs.
The survey of the Amarna hinterland is helping to expand our knowledge of how the city functioned and exercised control over the surrounding countryside, and it will aid in our understanding of how cities operated in Ancient Egypt in general. Without the recent development of high resolution GPS survey techniques, the recording of such a large landscape, with the degree of accuracy required, would have been a daunting task.
This work has been funded by the Egypt Exploration Society and was undertaken whilst the author was a research fellow with the Wetland Archaeology and Environments Research Centre, Department of Geography, University of Hull. The research has benefited from discussions in the field with Barry Kemp and Dr Corinna Rossi, but all the contents of the paper are fully the responsibility of the author. Special thanks must go to Dr Rossi in her capacity as snake scarer.
BRANTING, S. & G.D. SUMMERS. 2002. Modelling terrain: the Global Positioning System (GPS) survey at Kerkenes Dog, Turkey. Antiquity 76: 63940.
CHAPMAN, H.P. & H. FENWICK. 2002. Contextualising previous excavation: the implications of applying GPS survey and GIS modelling techniques to Watton Priory, East Yorkshire. Medieval Archaeology 46: 81-89.
CHAPMAN, H.P. & R. VAN DE NOORT. 2001. High-resolution wetland prospection, using GPS and GIS: landscape studies at Sutton Common (South Yorkshire), and Meare Village East (Somerset). Journal of Archaeological Science 28: 365-75.
DAVIES, N. DE G. 1906. The rock cut tombs of El Amarna Part 4: The tombs of Penthu, Mahu, and others. London: Egypt Exploration Fund, Archaeological Survey of Egypt Memoir 16.
FENWICK, H. 1999. Medieval moated sites in the Vale of York: distribution, modelling and wetland potential, in R. Van de Noort & S. Ellis (eds) Wetland Heritage if the Vale of York: an archaeological survey: 255-268. Hull: Humber Wetlands Project, University of Hull.
--2000. Medieval sites in the Hull valley: distribution and modelling, in R. Van de Noort & S. Ellis (eds) Wetland Heritage of the Hull Valley: an archaeological survey: 183-91. Hull: Humber Wetlands Project, University of Hull.
HEYWOOD, I., S. CORNELIUS & S. CARVER 2002. An Introduction to Geographical Information Systems. Harlow: Prentice Hall.
KEMP, B. 1989. Ancient Egypt: anatomy of a civilisation. London: Routledge.
KEMP, B. & S. GARFI. 1993. A survey of the Ancient City of El-'Amarna. London: Egypt Exploration Society Occasional Publications 9.
LEICK, A. 1995. GPS Satellite surveying. New York: Wiley.
PETRIE, W.M.F. 1894. Tell El Amarna. London: Methuen & Co.
TIMME, P. 1917. Tell El Amarna vor der Deutschen Ausgrabung in Jahre 1911. Berlin.
Helen Fenwick, * Department of History, University of Hull, HU6 7RX UK(Email:firstname.lastname@example.org)…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Ancient Roads and GPS Survey: Modelling the Amarna Plain. Contributors: Fenwick, Helen - Author. Journal title: Antiquity. Volume: 78. Issue: 302 Publication date: December 2004. Page number: 880+. © 2008 Antiquity Publications, Ltd. COPYRIGHT 2004 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.