Access to water is one of the greatest global challenges of the 21st century. Scholars from different fields of research around the world are dealing with the ever-growing demand for, and with the severe supply constraints of water. Ancient societies dealt with similar problems.

Within these constraints, a massive transformation in the organisation of water resources, agricultural systems and settlement patterns as well as communication networks allowed for an active exploitation of the Udhruḥ region, 12 km east of Petra (Southern Jordan), turning the steppe into green oases. After five years of prospective archaeological fieldwork (2011-2015) we could conclude that the research area around the town of Udhruḥ is one of the most complete and best preserved field ‘laboratories’ in the south of Jordan to study the long-term development of innovative water management and agricultural systems from Nabataean to Mamluk periods (1st century BCE – 15th century CE). We thus can distinguish three ancient agro-hydrological systems in the Udhruḥ region (figure 1). Systems that extract water from different sources and for different purposes, whereby agricultural use seems to be the prevailing aim. This research program focuses on the diachronic reconstruction of one of these antique agro-hydrological techniques – a subterranean qanat system with its associated surface structures and fields (figure 2-3) – which was employed to cultivate this arid landscape and the societal conditions that contextualised them. An international and interdisciplinary research team examines what the key to this water management and agricultural success was in ancient times.

Driessen2020 Fig1 2

Figure 1

Driessen2020 Fig2

Figure 2

Driessen2020 Fig3

For the 2019 field season we worked on the excavation and 3D-reconstructing several of the subterranean and surface structures of the Udhruh qanat system in the Wādī al-Fiqai. The monitoring of natural degradation processes of such structures retrieved in erosion gullies in this wādī by current flash floods is also part of our work and will be for the coming years. 35 scientific dating samples were taken of several parts of the system, of which a selection of twelve samples was made which were send for AMS/14C dating. Soil and water samples were extracted from the northern field systems connected to northern reservoir 1 in Wādī al-Fiqai. These samples are tested on fertility, biodiversity and remediation capacities for agricultural potential by soil scientists at Wageningen University (Department of Soil Quality - Chemical and Biological Soil Laboratory). The results show inter alia that these fields have been used for irrigated agriculture over a long period of time, and what the effects were for the involved soils.



Rapid population growths and changing climatological conditions, especially in some of the most water-scarce regions of the world, result in increasing pressures on already overexploited water resources. Such is especially the case for the arid and semi-arid parts of the Middle East, Northern and Sub-Sahara Africa, whereby the overwhelming majority of water is used for agricultural needs.[1] The annual precipitation in such regions is not only low and poorly distributed over crop-growing seasons, but a large part gets lost before becoming available for agricultural use. Different methods and techniques of water harvesting are employed to increase the availability of fresh water to agricultural crops in arid and semi-arid regions.[2]

Archaeological research makes clear that ancient societies were dealing with similar problems. In ancient times land-use systems and resource management, particularly elaborated water-harvesting schemes, were employed to prevent this loss.

The Udhruḥ Archaeological Project tries to shed insights on how people practised water procurement for agricultural purposes in the hinterland of Petra (South Jordan) in ancient times. After several years of archaeological field work we can already conclude that the research area around the village of Udhruḥ, 12 km to the east of Petra, is one of the most complete and best preserved field ‘laboratories’ to study the long-term development of innovative water management and agricultural systems from Nabataean till Mamluk times (1st century BCE – 15th century CE) in southern Jordan.


Archaeological research of ancient water harvesting and agricultural schemes

Three ancient agro-hydrological systems can be distinguished in the Udhruḥ region (figure 1). Systems that extract water from different sources and for different purposes, whereby agricultural use seems to be the prevailing aim.

These are:

  1. A perennial source in Udhruḥ was used to irrigate a patchwork of compound gardens, with retrieved ceramics that dates from the Nabataean period and onwards.
  2. In the hilly area northwest of Udhruḥ (Jabal ash-Sharah) a combination of ancient rainwater-catchment and run-off water harvesting techniques is observed, with surface finds and associated settlements dating from Nabataean and Byzantine times. These water harvesting schemes were employed to hold and direct run-off water that would otherwise get lost.
  3. An impressive network of well-preserved ancient subterranean and surface-water conservation measures and connected irrigated fields – a qanat-system - was recorded in a large flood plain largely covered by alluvial deposits southeast of Udhruḥ (Wādī al-Fiqai). The Udhruḥ qanat (figure 2-3) made use of the extraction of deep percolation water, whereby also solutions were applied to prevent loss through evaporation. This system seems to have been in use for longer periods of time. Intensive surveys on and around the fields provided us with a broad spectrum of material culture ranging from Nabataean, Roman, Byzantine, Umayyad and Ayyubid pottery to fragments of Mamluk glass bracelets. 


The Udhruḥ qanat and connecting fields[3]

The last system – the Udhruḥ qanat – will be our research focus for the involved program, because of the ingenuity and investments made on its construction and reworking, its long durée usage throughout several classical and Islamic periods, and its completeness in preservation. The antique agro-hydrological systems that transformed these parts of the steppe into green oases still seem to be largely intact and are (partly) buried by alluvial deposits (figure 3). Through a combination of contextual small-scale excavations and non-destructive ground-based methods (ground-penetrating radar, magnetometry, (aerial)photogrammetry, 3D-scanning) obtainable data will be retrieved on the position, dimensions and interconnections of these and other structures in this geologically diverse archaeological landscape. The comparative OSL and 14C dating will provide a solid chronological framework for inter alia the construction, maintenance and disuse of the Udhruḥ qanat-system and to the study of antique qanat-systems in general, but will also contribute to the refinement of these dating methodologies.

Trying to unravel the exact modus operandi of this qanat scheme and the continuity in use will be a great challenge for our research project in the coming years. Only by practicing an interdisciplinary approach can this be accomplished whereby archaeological research is integrated with historical, geophysical, water resources, bio-scientific and biogeochemical soil studies. On the one hand we think that this approach will help us to reconstruct the agro-hydrological landscape and the longe durée-development of this for the Udhruḥ region. On the other hand we hope that the interdisciplinary approach – applicant is both an archaeologist as an agronomist – will not only result in the comprehension of the antique semi-arid landscape management from a diachronic perspective, but will also lead to translational and innovative thinking which can contribute to possible sustainable agricultural and water management solutions for future use in these regions.


Report Antique Green Desert in the Udhruḥ Region Field Campaign 2019

The following activities have been employed at the qanat-system southeast of Udhruḥ in the Wādī al-Fiqai (figure 4) during the 2019 field campaign:

  1. Excavating two qanat shafts which will probably lead to the (ancient) water resources of the system.
  2. Cleaning and excavating SW corner of southern large antique water reservoir 2 in Wādī al-Fiqai, of which looting destruction with a bulldozer was observed during 2018 field campaign.
  3. Excavating several parts of the surface water management system in the Wādī al-Fiqai, measuring these with RTK GNSS GPS-Rover and making 3D-reconstructions of these.
  4. Collecting samples for 14C and OSL-dating. These are carried out by laboratories of University Groningen (Centre for Isotope Research) and Wageningen University (Netherlands Centre for Luminescence Dating).
  5. Collecting soil and water samples from the northern field systems connected to northern reservoir 1 in Wādī al-Fiqai. These samples are tested on fertility, biodiversity and remediation capacities for agricultural potential by soil scientists at Wageningen University (Department of Soil Quality - Chemical and Biological Soil Laboratory).

Driessen2020 Fig4

Figure 4. Scheme of 2019 field work in the Wādī al-Fiqai, southeast of Udhruḥ.


1. Excavating two qanat shafts.

These will probably lead to the (ancient) water resources of the system and provide us information on the layout and depth of the qanat shafts and horizontal channels, dating of its construction, maintenances. This will also provide knowledge on the getting out of use of the qanat system, and the possible aquiferous capacities of the geological layers it was constructed in. We started with excavating the most western qanat shaft near the ‘mother well’ (ground water level or aquiferous layer). After getting through a thick desert pavement the shaft was excavated for approximately 3m till the bedrock was reached where the qanat shaft was further cut out. Because of security reasons it was decided to make safety precautions. We hired a mechanical excavator to remove parts of the desert pavement and prepare the location – by removing part of the 3m thick Pleistone gravel layer on top of the bedrock – for safe excavation this qanat shaft during future field campaigns (figure 5). The formed trench has been demarcated by a fence and warning signs carried out by our team in order to prevent local shepherds from falling in.

Driessen2020 Fig5

Figure 5. Preparing the location of the most western qanat shaft by means of mechanical excavator.

We also started to excavate the most eastern qanat shaft near the outlet of the system. The location was based on the external characteristics of observed 231 qanat shafts in the area southeast of Udhruḥ. The excavation trench was excavated till 4m deep (figure 6). We will continue working with this in coming field season. No security preparations needed to be made here as the natural layers of Pleisocene gravel and old alluvial deposits are very hard and stable at this location. A fence with warning signs has been made here as well.

Driessen2020 Fig6

Figure 6. Excavating the most eastern qanat shaft.

2. Southern Reservoir 2: cleaning and excavating after looting damage.

The cleaning and excavating the southwestern corner of a large antique water reservoir (reservoir 2) was executed in the Wādī al-Fiqai, because of serious destruction with bulldozer by looters/treasure hunters as was observed during the 2018 field campaign. This work has been carried out and resulted in important information on the construction of this most southern reservoir, which has been laid out on Muwaqqar Chalk Marl bedrock (Maastrichtian Paleocene) and was partially carved out through a Pleistocene Gravel formation. The side wall of the reservoir consisted in a more than 2m wide ashlar wall, see figure 7.

Driessen2020 Fig7

Figure 7. The looted SW corner of southern reservoir 2 after cleaning.

This most southern reservoir (no. 2) had a capacity – measuring 33 x 34 x 2.7m – of at least 2,7 million liters of water. The mortar/concrete lining of the reservoir is 3cm thick from which we extracted charred remains for dating purposes (see point 4). It is expected that a coverage was employed to prevent it from serious evaporation, although no remains of such a structure has been discovered as yet. However also other means as covering the water surface with for instance organic mats of braided thatch or large leaves could prevent from such evaporation.[4] Everything was measured by means of RTK GNSS GPS-Rover. 3D reconstructions are made of all excavated parts of the system by means of photogrammetry and 3D-scanning, for the preliminary 3D reconstruction of this part of reservoir 2 see figure 8.

Driessen20 Fig8 2

Figure 8. Preliminary 3D reconstruction of SW corner of reservoir 2, which was looted in 2017/18.

3. Surface structures water management system Wad el-Fiqay.

The third operation employed during the 2019 field work was excavating several parts of the surface water management system in the Wādī al-Fiqai, measuring them with RTK GNSS GPS-Rover and making 3D-reconstructions of these. The interior southeast corner of northern reservoir 1 (connected to fields 1-2 sampled for soil fertility, see point 4) was excavated and cleaned as were parts of interior floor of reservoir (figure 9). This provided us with important information on the construction and use of this 50x50m reservoir with a capacity of at least 3.3 million liters of water. The differences in height between the western and eastern side of the reservoir were approx. 0.5m, enabling the reservoir to drain towards the agricultural fields in event of low water levels. Samples of charred remains of the applied mortar of this corner of reservoir 1, but also of other parts were taken in order to 14C date the construction of this reservoir, see also point 4. All features of the reservoir were measured with RTK GNSS GPS-Rover and 3D reconstructions of this corner have been made.

Driessen2020 Fig9

Figure 9. Excavation of SE corner of northern reservoir 1

We have also made a trial trench on the channel providing the northern reservoir 1 with water. In the field this could only be observed by a long line of piled stones. The trial trench made clear that this was a solid walled structure, made of a dry-stack wall, which has collapsed throughout time. On the lateral sides of this walled structure antique applied mortar could still be observed. Most probably the interior of this walled structure was composed of so-called adobe (pisé/Stamphlehm/daub) technique. Remnants of the water channel on top of this wall – made of mortar – could be observed in trial tranches on both sides of the dry-stack wall (figure 10).

Driessen2020 Fig10

Figure 10. Excavation of channel providing northern reservoir 1 with water.

Our hypothesis is that when the system went out of use the upper mortared part degenerated and the elements got hold on the structure which resulted finally in its collapse. Samples of charred remains of the applied mortar of this walled structure were taken in order to 14C date the construction of this channel, see also point 4. As with other remnants also a 3D reconstruction of this channel has been made. This technique of building with dry-stack walls and daub core with on top mortar-lined channels can be observed in the low-risk flooding zones of the system.

Another more durable technique of solid concrete based ashlar walls with on top mortar-lined channels has been applied in wadis and high-risk flooding zones (figure 11).

Driessen2020 Fig11

Figure 11. Solid water channel in high risk flooding zone of Wādī al-Fiqai.

We have also excavated, GPS measured and photographed all exposed water conduit structures in a very large erosion gully (> 10m wide, > 3m high and more than a kilometer long) observed in the Wādī al-Fiqai in 2011. According to dr. Fawzi Abudanah this erosion gully was not here during his PhD-research in 2004/6. We have made 3D reconstructions (figure 12) of this erosion gully. This has been done for direct research purposes, but also – from archaeological research and heritage management perspectives – to monitor future natural formation processes and possible natural degradations of the structures retrieved in this gully.

Driessen2020 Fig12

Figure 12. Preliminary 3D reconstruction of part of the erosion gully with retrieved antique water conduit structures


4. Collecting samples for 14C and OSL-dating.

In previous years (2014-2016) several samples were already analyzed (figure 16) by the laboratories of the Centre for Isotope Research (University Groningen – NL) and the Netherlands Centre for Luminescence Dating (Wageningen University – NL). During last year field work 35 new samples were collected for 14C dating purposes. These were extracted with specialized tools under non-organic contamination conditions from especially the interior mortar cores of different surface water management structures from Wādī al-Fiqai (see table 2). It was decided to extract predominantly charred twigs – most probably from the lime burning process for mortar production – from these mortars. This choice was made – and no samples of larger chunks of charcoals were taken – to avoid earlier dating as result of possible reuse or extended use of e.g. larger wooden architectural elements. At each location several samples were taken, sometimes of different structures and/or elements. A selection of 12 samples (find numbers 2624, 2649, 2661, 2674, 2676, 2678, 2679, 2684, 2685, 2686, 2688 and 2689, see table 2) was made and send for 14C dating by means of AMS to the Centre for Isotope Research (University Groningen). In December 2019 I received a message from dr. Sanne Palstra that due to the long waiting lists for 14C analysis, and the special preparation and arrangements for our small samples for AMS it will take another few months processing time before we can expect the final results of all samples. We will inform once the results of these have arrived.

The dating of the structures is an ongoing process which can only be secured when more dating samples have been analysed in order to obtain a more robust diachronic framework. This most probably will take some more years of research.

Driessen2020 table2

Table 2. 2019 dating samples for 14C from structures in Wādī al-Fiqai.

Driessen2020 Fig16

Figure 16. Earlier dating results of structures in Wādī al-Fiqai, plus locations of samples extracted in 2019 (last ones indicated by arrows and number of samples in red).

5. Collecting soil and water samples for soil capacity testing.

Twelve trial trenches were dug to collect soil samples as well inside the perimeters of ancient fields 1-2 connected to northern reservoir 1 (6 trenches), as well outside these fields for referencing (6 trenches; see figure 13). These samples were tested on fertility, biodiversity and remediation capacities for agricultural potential by soil scientists at Wageningen University (Department of Soil Quality - Chemical and Biological Soil Laboratory). We have taken 104 samples (63 soil samples in plastic bags, plus 41 soil samples in metal rings) which were analysed and processed by Ángel Velasco Sánchez MSc under surveillance of Marcel Hoosbeek at the Soil Chemistry and Chemical Soil Quality Centre of Wageningen University and Research Centre (Netherlands) for soil chemistry, physics and water carrying capacities. Water samples were collected, to be extracted from the infield soil samples, as well from a borehole pump near the ancient mother well of the qanat system. This was draining water from the current ground water level at -155m below surface level.

Driessen2020 Fig13

Figure 13. Soil samples were collected from 6 trenches inside the field system (blue) plus 6 trenches outside the field system (red).

The research questions for the antique agricultural activities by the soil science approach were twofold:

  1. Are the signs of ancient agriculture still measurable today?
  2. What kind of agricultural potential did these fields have?

For this the following lab analyses were executed on the soil samples:

  • Preparation (plus determination dry bulk density and water holding capicity) of each soil sample
  • Electrical Conductivity (EC = indirect way of measuring salt concentrations in the soil) and pH determination per soil sample
  • Cation Exchange Capacity (CEC = a measure of soil fertility, as it indicates the capacity of the soil to retain several nutrients (e.g. K+, NH4+, Ca2+) in plant-available form.
  • Available phosphorous (P) measuring, by the P-Olsen method.
  • Organic carbon content measuring by
  1. A) Loss on ignition method (LOI)
  2. B) Kurmies Method: Organic carbon content was measured by colorimetry on a spectrophotometer (Thermo Spectronic Aquamate).
  • Granulometry (particle size) measurements were carried out by laser diffraction on all samples
  • Multi trace element analysis primarily on water samples with an inductively coupled plasma mass spectrometry (ICP-MS) → resulting in peaks of isotopic forms of Lanthanum (La), Boron (B), Antimony (Sb), Uranium (U), Scandium (Sc), Lithium (Li), Titanium (Ti), Vanadium (V), Chromium (Cr), Manganese (Mn), Nickel (Ni), Molybdenum (Mo), Barium (Ba), Tungsten (W), Palladium (Pd), Cesium (Cs), Cobalt (Co), Copper (Cu), Tin (Sn), Indium (In), Germanium (Ge), Gallium (Ga), Strontium (Sr), Iodine (Y) and Rhodium (Rh).
  • The soil samples are also analysed and measured for these trace elements.

January 2020 the results came through.

For answering the first research question – are the signs of ancient agriculture still measurable today? – the first focus was on determining the effects of antique irrigation inside the fields. Irrigation will lead to the leaching out of mobile salts. There is evidence that this took place: there are significant differences in EC, Na and SAR (Sodium Adsorption Ratio) between samples from inside and outside the antique fields (see figure 14).

Lower amounts of LOI (organic C) are aIso indicative for potential salt leaching by irrigation water use (figure 14).

Significant differences in concentrations of mobile trace elements as e.g. Boron (B), Lithium (Li) and Strontium (Sr) are indicative as well for leaching by irrigation water use (figure 15).

Driessen2020 Fig14 1

Driessen2020 Fig14 2

Figure 14. Results of soil chemical and physical parameters analyses across the complete soil profile.

Driessen2020 Fig15

Figure 15. Increase of EC and second component / mobile trace element concentrations with increasing depth. EC (μS/cm) and trace elements (μg/kg).

Driessen2020 tab1

Table 1. Electrical conductivity, pH and trace elements measured on water samples. EC (μS/cm) and trace elements (μg/L).

Irrigation water – when not saline – can wash nutrients as trace elements from soil. The analysed water samples had a very low EC → very low salt concentration (table 1). This water contained relative high concentrations of immobile trace elements like Scandium (Sc) and Germanium (Ge). These were also found inside the fields – not outside the fields – resulting in the hypothesis that these ended up in the fields via the irrigation water.

Secondly also the plant uptake of the grown crops could result in the lower concentrations of the mobile trace elements as Lithium, Boron and Strontium inside the fields versus the surroundings. Thirdly also the practised tillage could contribute to the depletion of these elements in soils of ancient fields as well. Altogether it however it becomes clear that the ancient fields have been used intensively for agricultural purposesif compared to the surrounding areas.

In order to answer the second research question – which kind of agricultural potential did these fields have? – we need many more samples to be taken from other antique fields and their surroundings, but also from other soils from the region. This will be executed throughout the 2021-2022 field campaigns. What already becomes clear is that the soils are alkaloid and have a rather high pH (figure 14). This makes them not very suitable for cereal and vegetable growing, but such soils can be quite appropriate for special vines and other fruits.

Dr. Mark Driessen
Faculty of Archaeology
Leiden University

Dr. Fawzi Abudanah
Al-Hussein Bin Talal University
Petra College for Tourism and Archaeology





[1] See Gleick 2014, 227-235; Data Table 2: Freshwater Withdrawal by Country and Sector.

[2] Oweis et al. 2012, 31-71.

[3] See Driessen/Abudanah 2018 for the composition and details of the Udhruḥ qanat gathered through earlier years of research.

[4] Driessen/Abudanah 2018, 145 and footnote 9.