t: +20 12 5113357 (mobile)
t: +20 27955666 (office)
False-colour Landsat satellite image of the Nile Valley, Amarna slightly left of centre. North is to the left.

The Panehsy Church Project, 2006

Contents

Interpreting past landscapes and project overview
Method
Results
Great Temple enclosure wall
Conclusion

Interpreting past landscapes and project overview 

Landscape reconstruction is a crucial issue within the field of Egyptian archaeology: so much of Egypt’s modern landscape continues to influence how we perceive the overall ancient landscape. It is difficult to step back and let go of preconceived notions of land levels, Nile flood patterns and ancient settlement locations. Ancient Egyptian history is, after all, embedded in its landscapes, and how we have interpreted key datasets is ultimately bound within our understanding of the Nile, desert and floodplains in the ancient Egyptian mindset. Ultimately, we must rely on a variety of archaeological, technological and geomorphological tools in reconstructing past landscapes.
 
The 2006 Middle Egypt Survey Project took place from March 29th to April 11th at Tell el-Amarna, with the intent of conducting coring on the east bank in the cultivation next to the ancient city. Overall project goals included finding evidence for any remains of Akhetaten beneath the current cultivation levels, locating evidence for the ancient Nile river course, and determining landscape evolution over the past 3300 years. The project cored at 30 sites in various locations in the floodplain directly to the west of Akhetaten. In addition, the project utilized a Quickbird satellite image (60 cm pixel resolution) taken in March of 2004 in order to locate potential subsurface archaeological remains at Amarna.

The bailer/boring equipment
The bailer/boring equipment

The crew using the coring equipment
The crew using the coring equipment

Method

This project utilized an Eijelkamp bailer/boring set, complete with core bits for use in sandy, silty and dense clay soil levels. The core pieces can reach a depth of 7 m, with either a 7 cm or 10 cm diameter. Core bits are changed when different soil horizons are reached. The floodplain next to Amarna necessitated frequent changing of core bits, in particular in areas of dense Nile clays. At a constant depth across the cored areas (taking into account the natural slope of the land), the core would hit dense desert gebel.

It is necessary to work with a trained crew, as it can be easy to lose core bits. Reis Omer Farouk Sharid and Reis Said Mohammed ably assisted with leading the crew of local men, and permitted the team to core an average of 10–12 m in a single day.

Recording and bagging the soil samples
Recording and bagging the soil samples

Sieving the soil samples
Sieving the soil samples

Recording each core level’s texture, nature (water laid, wind laid or artificial), grain size, structure, consistency, color, depth, and inclusions is crucial for overall core interpretation. Each core is then individually bagged for sieving. Sieving at 1 cm, 5 mm and 1 mm allows for smaller pieces of pottery, stones, charcoal, bones, metal and glass to be collected for further analysis, and permits the evaluation of each soil type.

Slide/analysis of material culture
Slide/analysis of material culture

Differing soil types indicate the development of the landscape, and may show former flood depositional levels. They also may indicate different occupation or cultivation horizons. After sieving, analyzing the material culture within each core sample allowed for dating each core. For example, the upper 0.5 m of each core would often contain modern debris. Below 0.5 m, a combination of Late Roman Period and Amarna Period material appeared, depending on the depth and overall disturbance of the earth. Treating each core as an archaeological ‘spit’ level permitted the comparison between different cores taken along similar points. 

Coring taking place
Coring taking place

Each day, the project team took between two and eight cores, depending on the nature of the soil. Taking cores at set intervals within the floodplain running north to south meant that a full range of results could be obtained. The total cores the project took numbered thirty, which ranged from 70 cm to 7 m in depth. Through the use of a GPS unit, the project recorded the Easting and Northings of each coring point for placement on a Quickbird satellite image of the entire Tell el-Amarna region for comparative purposes.
Initially, the project tested three cores in the floodplain area directly to the west of the southern extents of the central city, followed by a series of cores in Kom el-Nana, to the North of El-Hawata and in El-Hagg Qandil. The rest of the cores followed a fairly set transect from the North City running south to the North Suburb, North Palace, Central City and South Suburbs.

The results of Ian Mathieson’s resistivity work of 1987, as reported in Amarna Reports V, 150–1 The results of Ian Mathieson’s resistivity work of 1987, as reported in Amarna Reports V, 150–1

Results

A major question to be examined is to what extent resistivity work can help find the ancient edge of Amarna in comparison with the coring and the satellite remote sensing work. Ian Mathieson conducted two resistivity seasons at Amarna in the mid 1980s, done from the Great Palace eastwards to determine course of the river, ancient ruins, and the slope of desert topography. Mathieson felt that while modern cultivation might have cut into raised floor levels of the ancient buildings, the original ground level would be beneath modern fields. He made 48 traverses, which he said suggested that archaeological debris existed at a depth of six metres, and said the resistivity suggested the former watercourse, which can be seen in this image. The coring work suggests a different interpretation of the landscape geomorphology.

Measured and drawn section of a core Measured and drawn section of a core

The overall project results revealed several key points about the geomorphology of the Amarna region, and show the extent to which the ancient city remains beneath the current cultivation levels. Overall, the cores show a variety of sloping silt, sand and clay layers beneath which lies the desert gebel on which Akhetaten was built. To the north of the North Palace, in the ‘harbor’ area, cores of 7 m revealed the remains of a likely waterfront location, as gebel was found at this depth in all other cores and not found here.  This may show a former part of the ancient Nile river course that has since shifted to the west, or, more likely may reveal a channel dug from the Nile (which as a whole seems to have shifted to the east in the last 2000+ years). 

At the village of Et-Till, the project made a deep core (6.94 m), while a well dug to 7 m revealed similar archaeological and geomorphological patterns. The core showed that the Late Roman Period landscape built up the area around Et-Till on top of the desert gebel, providing fertile soil for cultivation, and suggests the central part of a major LRP settlement closely connected to the Christian establishments documented in the area, one being at the North Tombs. A distinct lack of Nile silt or clay above the gebel revealed that the Nile flood levels did not reach the edge of Akhetaten until the Late Roman Period, which may be when a large part of the city was simply washed away.

No Amarna Period material is apparent in the cores to the south until the Great Palace area. Within the Central City area, several deep cores showed limited Amarna Period pottery, with Late Roman Period pottery appearing in the upper 2 m of each core. Confirmed non-Late Roman Period material, likely New Kingdom sherds based on wares, appears in the cores taken in and to the west of El-Hagg Qandil, where a definite post-Amarna Period settlement exists beneath the town.

Little of the ancient city remains beneath the current floodplain level, which appears to have developed in the past 2000 years, to judge from the concentration of Late Roman Period sherds in the silt layers some 2–5 m below the ground. ‘Modern’ (i.e. post Late Roman Period) cultivation and the Nile flood levels appear to have combined to obliterate any remaining traces of the city, save a few areas where deep foundations preserved Amarna Period material culture.

Great Temple enclosure wall

The enclosure wall of the Great Aten temple, as seen from the Quickbird satellite. The northern part of the enclosure is covered by a modern cemetery
The enclosure wall of the Great Aten temple, as seen from the Quickbird satellite. The northern part of the enclosure is covered by a modern cemetery

Using Quickbird’s multispectral capabilities, we can gain a clearer picture of what lies beneath the soil. Its multispectral bands allow for subsurface mud brick to be differentiated from the surrounding sands, due to their higher moisture content. For the first time, we can use satellite imagery as an airborne resistivity survey tool, detecting subsurface structures and enabling more efficient mapping of sites. Even parts of Amarna that have been built over, such as the northern enclosure wall surrounding the Great Temple, are still visible with Quickbird imagery.

Conclusion

The coring work during the 2006 season of the Middle Egypt Survey Project has assisted with answering both geomorphological and archaeological questions, but like any work of this nature, it has raised many questions which future coring work will aim to address, including locating former Nile courses to see if the river was actually located a significant distance to the west, as well as using additional deep coring, in some cases 10+m, to locate older archaeological sites on the west bank. This information will, in turn, be used to locate the ancient Nile river course(s) in relationship to Akhetaten’s actual ancient borders.

Author: Sarah Parcak (University of Birmingham, Alabama)

Satellite Image of
Amarna

 
 

Website first posted September 2000; last updated November 2010 | enquiries concerning website: email bjk2@cam.ac.uk