From the End of the Pleistocene to about 5000 years ago, much of the Sahara is thought to have been covered by a lush green savannah, with numerous, now dry, lakes and rivers, providing a home to large Human and Animal populations. During this time the Nile was a much more substantial waterway, wider and deeper than it is today, prone to major flood events, and surrounded by extensive marshes and wetlands, making the Nile Valley a difficult environment for the hunter-gatherers of the period, and it appears to have been largely uninhabited.
From around 5500 years ago the climate of the Sahara began to dry out, with the lush grasslands that had covered much of the area slowly disappearing, while the Nile Valley became a habitable refugia, attracting people to settle in the area, settle and begin farming. This environment enabled a great cultural flourishing, and eventually the formation of one of the word's first centralized states, the Old Kingdom of Egypt, which formed around 2686 BC. The Old Kingdom quickly established a culture of monumental agriculture, which most notably manifested as the great pyramids of the period.
At this time the discharge of the Nile was still much higher than it is today, and multiple branches of the river are thought to have been present at any time, meandering across the floodplains of Sudan and Egypt. Since this time the Nile Valley in Egypt has undergone significant changes, due to the decline in water flow over the past four-and-a-half millennia, and the building of the Aswan High Dam in the 1960s, which made the Nile Valley in Egypt a much more controlled and predictable environment.
To the Ancient Egyptians the Nile was a vital resource, providing regular floods which kept the fields irrigated and fertile, and a ready-made transport network which enabled them to move food, manufactured goods, and building materials, to wherever they were needed. However, they had much less control of this resource than modern Egyptians, and it was prone to migrating laterally across the desert, leaving settlements and building projects cut off from water resources and easy transport. The location of many ancient sites, far from the current course of the river, provides a testament to the movement of the river, but the exact course of the river in ancient times is less clear, difficult to map due to modern farms and settlements covering the floodplain. Thus, we remain uncertain where the ancient courses of the river ran, nor if more than one branch was active at any given time.
The pyramids and great temples of the Old Kingdom, which logically would have been constructed close to a navigable waterway, area arranged in a roughly linear pattern, running parallel to the Nile from Lisht in the south to Giza in the north.
In a paper published in the journal Communications Earth and Environment on 16 May 2024, Eman Ghoneim of the Department of Earth and Ocean Sciences at the University of North Carolina Wilmington, Timothy Ralph of the School of Natural Sciences at Macquarie University, Suzanne Onstine of the Department of History at the University of Memphis, Raghda El-Behaedi of Near Eastern Languages and Civilizations at the University of Chicago, Gad El-Qady of the Egyptian National Research Institute of Astronomy and Geophysics, Amr Fahil of the Geology Department at Tanta University, Mahfooz Hafez, Magdy Atya, Mohamed Ebrahim, and Ashraf Khozym, also of the Egyptian National Research Institute of Astronomy and Geophysics, and Mohamed Fathy, also of the Geology Department at Tanta University, present the results of a study which used remote sensing, geomorphological, soil coring and geophysical evidence to map the Western Desert between Lisht and Giza, demonstrating the presence of a lost river branch running along this course.
Ghoneim et al. used Synthetic Aperture Radar to create a three dimensional map of the Nile Floodplain and the surrounding desert, finding evidence of a former river channel between Lisht and Giza, which would have been navigable during the Old Kingdom, and until the Second Intermediate Period, passing by 31 pyramids from the Third to Thirteenth Dynasties. They name this the Ahramat Branch, from the Arabic 'al'ahramat' (الأهرامات), meaning 'pyramids'. This branch is largely covered by fields on the modern Nile Floodplain, making it invisible in visual satellite images, but can be traced on the ground surface by TanDEM-X satellite radar imagary and Topographic Position Index functions on Geographical Information Systems software. This channel lies between 2.5 and 10.25 km to the west of the modern course of the Nile River. This channel appears to have been about 64 km in length, between 200 and 700 km wide, and between two and eight metres deep. The size of this waterway, combined with its course directly past the pyramids and other works of monumental architecture now marooned in the desert, implies that it was of great functional importance to the early Ancient Egyptians.
The water course of the ancient Ahramat Branch. (a) Shows the Ahramat Branch borders a large number of pyramids dating from the Old Kingdom to the 2nd Intermediate Period and spanning between Dynasties 3 and 13. (b) Shows Bahr el-Libeini Canal and remnant of abandoned channel visible in a 1911 historical map (Egyptian Survey Department scale 1:50,000). (c) Bahr el-Libeini Canal and the abandoned channel are overlain on satellite basemap. Bahr el-Libeini is possibly the last remnant of the Ahramat Branch before it migrated eastward. (d) A visible segment of the Ahramat Branch in TanDEM-X satellite radar imagary is now partially occupied by the modern Bahr el-Libeini Canal. (e) A major segment of the Ahramat Branch, approximately 20 km long and 0.5 km wide, can be traced in the floodplain along the Western Desert Plateau south of the town of Jirza. Location of (e) is marked in white a box in (a). Ghoneim et al. (2024).
A survey carried out using Ground Penetrating Radar and Electromagnetic Tomography found a former river channel was buried beneath 1-1.5 m of modern cultivated sediment, which matched the position of the channel observed in the satellite radar imagery. The sediments within this channel are different to both the overlying silt deposits of the modern floodplain and the sandy sediments on either side of the channel. At the site where the ground survey was carried out, the riverbed sediments were approximately 400 m wide and 25 m deep (the depth of riverbed sediments implies the presence of a river for a long period of time, rather than relating to the depth of the river).
Two sediment cores were taken in this area. The first of these, Core A, was taken between the centre of the channel and the left bank, and reached a depth of 13 m. This core had an upper layer of sandy brown mud which reached down to about 2.7 m. Beneath this was a layer containing limestone and chert fragments, a reddish sandy mud layer with gravel and handmade material inclusions, then from about 3 m to about 5.8 m, another reddish sandy mud layer with gravel and Freshwater Mussel shells, followed by another layer of limestone and chert fragments, a reddish sandy mud layer with gravel and handmade material inclusions, then a black sandy mud from 6 m to 6.8 m, which graded into a well sorted medium sand; this was clean of mud by about 8 m, and persisted till about 13 m. The second core, Core B, was taken on the right bank of the branch, and reached a depth of 20 m. Here, recent brown sandy muds reached a depth of 1.5, beneath which alternating brown and gray layers of silty and sandy mud down to about 4 m. This was followed by a black sandy mud layer from 4 m to 4.9 m, then another reddish sandy mud layer with gravel and freshwater mussel shells till about 5 m. From 5 m to 20 m the profile was dominated by a clean, well sorted, medium sand. In both cores the sand layers contained groundwater, suggesting that the former Ahramat Branch of the Nile still serves as a conduit for subsurface water flow within the Nile Floodplain.
Deep sediment cores from the southern segment of the Ahramat Branch. It shows two-soil cores, core A and core B, with soil profile descriptions, graphic core logs, sediment grain size charts, and example photographs. Ghoneim et al. (2024).
The pyramids of Ancient Egypt were not isolated buildings. Each formed part of a temple complex, which also included a mortuary temple, which stood next to the pyramid, and a valley temple, beside the waterway and some distance from the pyramid and mortuary temple. All of these were connected by causeway (raised walkway), which provided a path to the pyramid complex. Many of the causeways associated with pyramids on the Ahramat Branch have causeways arranged at right angles to the course of the waterway, and end where the riverbank would have been.
The valley temples of Ancient Egyptian pyramid complexes served as disembarkation points for visitors arriving by river. Many of these valley temples have never been found, and are presumed either to have been totally plundered for their stone or to be buried beneath the desert sands or agricultural lands of the modern Nile Valley. Five valley temples are still at least partially present on the Ahramat Branch. These are associated with the Fourth Dynasty Bent Pyramid, Pyramid of Khafre, and Pyramid of Menkaure, the Fifth Dynasty Pyramid of Sahure, and the Sixth Dynasty Pyramid of Pepi II, all of which date from the Old Kingdom. All of these valley temples are adjacent to the Ahramat Branch waterway, indicating that this branch of the Nile was a navigable waterway during the Old Kingdom.
Ghoneim et al. use the positions of the pyramids arranged along the Ahramat Branch to reconstruct a history of the waterway. They conclude that the waters of were highest during the Fourth Dynasty, with pyramids from this time placed on high elevations far from the waterway. The water appears to have been much lower during the Fifth Dynasty, with pyramids consequently placed at much lower elevations, closer to the floodplain. The waters appear to have begun to rise again during the reign of Unas, the last Pharaoh of the Fifth Dynasty, and to have continued to rise during the Sixth Dynasty, with the pyramids of Pepi II and Merenre located deep into the desert.
The pyramid of the Eighth Dynasty Qakare Ibi is again located low and close to the riverbed, suggesting that waters had again fallen by this time. Qakare Ibi lived during the First Intermediate Period, after the fall of the Old Kingdom, which is known to have been associated with a significant drought, during which the annual flood on the Nile failed to appear for between 30 and 40 years. Sediment cores takin in Memphis, the capital of the Old Kingdom, have shown the city to have been covered by about 3 m of windblown sands. The Ahramat Branch seems to have shifted significantly to the east during the First Intermediate Period, then continued to drift eastwards during the Middle Kingdom, before effectively drying up during the Second Intermediate Period.
Sentinal 1 radar images show a number of apparent channels running from the Western Desert Plateau to the Ahramat Branch, suggesting tributaries to the river. All of these are now covered by sand, and therefore invisible in spectrographic images, but appear dark in the radar images, suggesting fluvial sediments beneath thr surface. Notably, the Sentinal 1 radar system, which can penetrate about 50 cm of dry sediment, is incapable of seeing the main path of the Ahramat Branch, which is largely covered by floodplain farmland. The inlets were visible in TanDEM-X satellite radar imagary, which suggested an extensive network of waterways in this area.
Using Radar and the Topographic Position Index for mapping major channels (inlets) connected to the Ahramat Branch. (a) Conceptual sketch of the dependence of surface roughness on the sensor wavelength λ. (b) Expected backscatter characteristics in sandy desert areas with buried dry riverbeds. (c) Dry channels/inlets masked by desert sand in the Dahshur area. (d) The channels’ courses were extracted using Topographic Position Index. Negative Topographic Position Index values highlight the courses of the channels while positive Topographic Position Index signify their banks. Ghoneim et al. (2024).
Several pyramids from the Fourth and Sixth dynasties appear to have causeways that connect to these inlets, rather than the main Ahramat Branch waterway. Thes include the Bent Pyramid in the royal necropolis of Dahshur, the first of three pyramids built by Sneferu, the first Pharaoh of the Fourth Dynasty, and one of the largest and oldest pre-Giza pyramids. The location of this pyramid has long been a mystery, as it is deep within the Eastern Desert, far from the modern Nile Floodplain. The Bent Pyramid has an impressive 700 m limestone block causeway, connecting it to a valley temple, which until now has been mysteriously distant from any body of water, facing onto the dry Wadi al-Taflah. Ghoneim et al.'s reconstruction places the causeway on an inlet of the Ahramat Branch, which they name the Dahshur Inlet, and which would itself have been more than 200 m wide. Ghoneim et al. reason that this inlet would still have been flooded during the reign of Sneferu, and that it would have played a major role in the movement of the materials used to construct the temple complex. The inlet could also have been used to move materials for the construction of the Red Pyramid (Sneferu's second), which is also in the Dashur area, although no trace of a causeway or valley temple associated with this pyramid remain. Several Middle Kingdom pyramids are also found in this area, including the Black Pyramid of Amenemhat III, the White Pyramid of Amenemhat II, and the Pyramid of Senusret III, although these are all at least a kilometre to the east of the Old Kingdom pyramids, lending support to the idea that the Ahramat Branch migrated eastward over time.
The Sakkara and Dahshur inlets are connected to the Ahramat Branch. (a) The two inlets are presently covered by sand, thus invisible in optical satellite imagery. (b) Radar data, and (c) TanDEM-X topographic data reveal the riverbed of the Sakkara Inlet due to radar signals penetration capability in dry sand. (b) and (c) show the causeways of Pepi II and Merenre Pyramids, from Dynasty 6, leading to the Saqqara Inlet. The Valley Temple of Pepi II Pyramid overlooks the inlet riverbank, which indicates that the inlet, and thus Ahramat Branch, were active during Dynasty 6. (d) Radar data, and (e) TanDEM-X topographic data, reveal the riverbed of the Dahshur Inlet with the Bent Pyramid’s causeway of Dynasty 4 leading to the Inlet. The Valley Temple of the Bent Pyramid overlooks the riverbank of the Dahshur Inlet, which indicates that the inlet and the Ahramat Branch were active during Dynasty 4 of the Old Kingdom. Ghoneim et al. (2024).
The satellite radar data suggests another inlet is located about 6 km to the north of the Dashur and to the west of the ciry of Memphis, which Ghoneim et al. refer to as the Sakkara Inlet. This is interpreted as a broad river, which would have been 600 m wide where it met the Ahramat Branch. The causeways of the Sixth Dynasty pyramids of Pepi II and Merenre, both part of the Sakkara Necropolis, have causeways which lead to the Sakkara Inlet. The causeway of Pepi II's pyramid runs northeast from the pyramid, across the Sakkara Plataeu to reach a valley temple located on the south bank of the Sakkara Inlet. The causeway of the Merenre Pyramid runs 350 m southeast from the pyramid, reaching the north bank of the Sakkara Inlet. Together these pyramids indicate that the waters of the Ahramat Branch were high enough during the Sixth Dynasty to flood at least the lower parts of the western inlets, enabling their use as waterways for the shipping of building materials. However, no Fifth Dynansty pyramids are located on the inlets, suggesting that the waters were lower during this interval.
Ghoneim et al.'s data set also indicates the Fourth Dynasty pyramids of Khafre, Menkaure, and Khentkaus, located on the Giza Plateau, had causeways leading to a smaller, but clearly still significant, inlet of the Ahramat Branch, which they term the Giza Inlet. The Khufu Pyramid (the largest of the Egyptian pyramids), appears to have been connected to the main Ahramat Branch. The locations of these pyramids and their causeways offer further proof that the Ahramat Branch and its eastern inlets were active waterways during the Fourth Dynasty, and fits with previous studies which have found evidence for marshy or riverine environments on the floodplains to the east of the Giza Pyramids during the Old Kingdom.
TanDEM-X data shows, in 3D, a clear topographic expression of a segment of the former Ahramat Branch in the Nile floodplain in close proximity to the Giza Plateau. The causeways of the four Pyramids lead to an inlet, which is named the Giza Inlet, that connects from the west with the Ahramat Branch. These causeways connect the pyramids with valley temples which acted as river harbours in antiquity. These river segments are invisible in optical satellite imagery since they are masked by the cultivated lands of the Nile floodplain. The photo shows the valley temple of Khafre Pyramid. Ghoneim et al. (2024).
Ghoneim et al.'s findings suggest that water levels in the Ahramat Branch were high during the Old Kingdom of Egypt, and particularly high during the Fourth Dynasty, after which they began to decline. This correlates with previous studies which have found that the discharge from the Nile into the Mediterranean was particularly high during the Fourth Dynasty, but decreased significantly afterwards, with the lowest output of the entire Pharaonic period seen during the Fifth and Sixth dynasties. Water levels in the Ahramat Branch appear to have been much lower during the Middle Kingdom, with structures assumed to have been built close to water much further to the east than in the Old Kingdom. This is at odds with previous studies of the hydrology of the Middle Kingdom, which have suggested a predominantly wet climate, with sporadic severe droughts. Ghoneim et al. explain this discrepancy as being due to the Ahramat Branch migrating laterally to the east, rather than simply containing less water, as in the Fifth Dynasty, though with a similar result as the building of new structures migrated eastwards.
Ghoneim et al. suggest that the eastward migration and then failure of the Ahramat Branch of the Nile may have been due to tectonic activity, as the Nile Delta and floodplain slowly tilted to the northeast, causing the western part of the floodplain to become elevated. This would lead the Ahramat Branch to migrate to the east, and eventually fail as the upper portion of the Branch fused with the main Nile channel. This eastern movement would also lead to the western portion of the floodplain to become more arid, enabling the sands of the Western Desert to intrude into the area, which would in turn lead to a higher sedimentation rate within the river, and a tendency to silt up during intervals of lower flow.
During the First Intermediate Period, which separates the Old Kingdom from the Middle Kingdom, Egypt experienced a prolonged drought, with settlements which had been prosperous during the Old Kingdom being engulfed by sands from the Western Desert. The Ahramat Branch seems to have significantly silted up during this interval, causing it to migrate further eastwards when wetter conditions returned prior to the rise of the Middle Kingdom.
During the Old Kingdom the Ahramat Branch was still connected to a number of inlets from the west, which were likely remnants of active drainage systems in the Tertiary or Pleistocene, when the Sahara was green and rainfall abundant across the region. These inlets would most probably have been fed from the Ahramat Branch rather than have carried water into it by the Old Kingdom, providing sheltered areas where the water would have been predicably calm, ideal for unloading materials for major construction materials, but would have dried up as the climate become more arid and the river shifted to the east.
Both of the sediment cores show an abrupt shift from well sorted medium sands to a more variable system with layers of gravel, fine sand, shell and other material, which Ghoneim et al. interpret as a sign of a shift from a constant, high-energy system within a deep, strongly-flowing river, to a more variable environment with a typical weaker flow, but periodic flood events. Two cores with similar profiles were previously drilled close to Giza, where they were interpreted as evidence for a Late Holocene floodplain palaeo-environment, utilised by people and prone to flood events. Further coring is likely to determine more features of the Ahramat Branch and the wider Nile floodplain.
Ghoneim et al. suggest a connection between the Ahramat Branch and the Bahr Yusef, a heavily canalized waterway connecting the Fayum Oasis to the Nile. It is likely that this waterway originally flowed northward before turning west and flowing into the Fayum Depression; the waterway currently bends very sharply westwards into the depression, possibly as a result of Human interference. The British Egyptologist John Römer has suggested that during the Middle Kingdom a dam was constructed which prevented the Bahr Yusef from flowing further to the north when its sluices were closed. Such a structure would have enabled the Egyptians of the Middle Kingdom to control the flow of the Ahramat Branch, or store water in the Fayum Depression.
Ghoneim et al. believe their evidence demonstrates the presence of a major former channel of the Nile, which they term the Ahramat Branch, which ran past the foothills of the Western Desert Plateau, the area where the majority of the Ancient Egyptian pyramids are found. The scale of this waterway, its proximity to the pyramids, and the numerous pyramid causeways which terminate at its banks, all strongly imply the branch was active during the Old and Middle Kingdoms, and was used as a navigable waterway, used to shift materials by the people of the period. Such a waterway would have provided an important transport link to the Ancient Egyptians, connecting towns and religious centres which have hitherto appeared disconnected. The eastward migration and eventual disappearance of this waterway, was driven by a combination of tectonic movements and the incursion of wind-blown sand from the Western Desert, combined with increasing desertification within the Great Sahara region, and the Nile Valley towards the end of the Old Kingdom.
The study shows the value of combining data from satellite imagary with drilled cores, the combination of which enabled the discovery of a significant waterway, which would have played a major role in the lives of the ancient inhabitants of the region.
This discovery will enable archaeologists to gain a much better understanding of the placement of settlements and temple complexes in Ancient Egypt, potentially leading to the discovery of new sites, and providing an opportunity to protect these sites from spreading urbanization in the region.
See also...
