Following J. C. Wilkinson, qanatFootnote 1 technology involves the exploitation of groundwater by galleries that lead to the surface. The water is taken from the water table and, thanks to a gentle slope and the help of gravity, is carried to a corresponding settlement, which can be several kilometers away. Therefore, qanat technology does not depend on the rhythm of natural rainfalls. The flow of water is continuous in the underground galleries, although it can vary seasonally in some areas. The gradient of the galleries is generally between 1 and 3‰ (Boucharlat 2017a, p. 282) and their length can reach several kilometers, with access shafts regularly spaced along their course.

This hydraulic system is found in many places across the Near and Middle East, like Saudi Arabia,Footnote 2 Yemen,Footnote 3 Central Asia, and Iran, but also in Spain (in Madrid this system is called viajes de agua). One of the most famous examples from Madrid, located in the city center, was excavated several years ago. Visitors to, and residents of Madrid, can visit this qanat, dating to the 16th century AD, using the subway at Opera train station close to the Royal Palace. The Royal Village of Aranjuez, located only 50 km south of Madrid, also incorporated a net of viajes de agua that have been studied since 1775.Footnote 4 However, this paper focuses on the Oman Peninsula, where qanats are called falaj (pl. aflāj).

In the United Arab Emirates, there are about 30 known aflāj, some of them still in use, while in Oman more than 70% of the water consumed in the early eighties came from aflāj, and 55% of irrigated crops took water from this system. These data are not extraordinary if we consider that the government has cataloged 44.066 aflāj in Oman, 3.023 of which are still in use.Footnote 5

In the Oman Peninsula, many aflāj are located in areas with a very high water table (like in the al Madam Plain which is situated in the interior of the Oman Peninsula and belongs in its totality to the emirate of Sharjah, UAE), or in regions with large concentrations of wadiān that allow water to seep into the ground (like the Batinah coast, Oman).

Qanat/falaj or catchment gallery

Since the 1990s, archaeological teams working on Iron Age settlements (i.e. Hili, Qarnt Bint Sa’ud, Maysar, Muweilah) have been contemplating the relationship between these sites and qanats; according to researchers, the development of these structures, which accessed underground water resources, would explain the establishment of populations on the fringe of the desert, and the growth of villages during this period (Fig. 1). Indeed, overall settlement numbers increased during the Iron Age, and these settlements were spread more widely across the region than ever before. This seems to be related to the diversification of resources and environments exploited during this period (Magee 2000).

Fig. 1
figure 1

Oman Peninsula showing the most important sites from Iron Age. Mañé (2005, p. 263)

Full size image

The oldest aflāj in the Oman Peninsula can be dated to the Iron Age (al Tikriti 2011). Unfortunately, we know little about the nature of the groundwater source tapped by these early structures. It has been suggested that the water came from alluvial fans or wadi underflows. These water sources would be sensitive to climate change (Boucharlat 2003, p. 162). Therefore, it is assumed that the abandonment of several aflāj at the end of the Iron Age was related to aridification and the subsequent depletion of shallow groundwater resources.

If the water table drops, it is necessary to re-excavate a falaj along its whole length until water once again flows into the gallery. If this occurs, then the old tunnel slope is no longer sufficient; for this reason, it must be corrected and the outflow of the falaj has to be moved downstream. As a consequence, water can no longer reach the original fields and settlements. There are various solutions to this problem:

  • Fields and settlements are moved downstream, near the new outflow of the falaj, as observed in Maysar 46Footnote 6 (Oman)

  • Fields are lowered and sunk into the ground in order to reach the level of the outflow.Footnote 7 The excavated soil is stacked at the edge of the oases, giving rise to large mounds called nadds, as can be seen at Nad Ziba (Ras al Khaimah)Footnote 8

  • A large reservoir is constructed into which the water flows. This case is archeologically documented in Qarn bint Sa’ūdFootnote 9 (Abu Dhabi) and nowadays at HattaFootnote 10 (Dubai)

Al Madam

The area of al Madam-Mleiha, which lies northwest of the Oman Peninsula, is a steppe sprinkled with trees and dotted with cultivated fields and palm groves that have been irrigated by wells and aflāj from Pre-Islamic times up to the present day (Fig. 2). The present inhabitants of this region are exploiting a water table that nowadays lies considerably deeper than in Antiquity. Water shortage, and its impact on the evolution of settlement patterns, has for a long time captured the attention of scholars working on the Iron Age of the Oman Peninsula. Evidence from al Madam provides new insight on this process. The Spanish team concentrated its work on sectors 1 (the Iron Age II-III settlement)Footnote 11 and 2 (falaj).Footnote 12 We established that al Madam was supplied with water coming from wells (Fig. 3) and water-draining galleries, as early as Iron Age I. A falaj system seems to have been the best choice for providing water in this context because of low rates of evaporation and the fact that the water table was shallower in antiquity. Precipitation would not have been sufficient for dry farming since the average annual rainfall did not exceed 100 mm and would have been highly variable from year to year. Water capture in the area (by wells or falaj) must have been sizeable if we note that Thuqeibah (the name of area where our main excavation is located) means small well, small pool, pit, or can even mean “small mouth of a falaj”.

Fig. 2
figure 2

Map showing the inland basin (Meliha-al Madam region) with major archaeological sites. Jasim et al. (2016, p. 11)

Full size image
Fig. 3
figure 3

Inner stratigraphy of al Madam 1-Thuqeibah well. (M. Nuñez Villanueva, Spanish Archeological Team)

Full size image

When archaeological surveys began in the al Madam area, Iraqi scholars reported the remains of a mudbrick village in the modern hamlet of Thuqeibah. This site, referred as al Madam 1-Thuqeibah,Footnote 13 revealed itself as the first stable settlement in the region at the beginning of 1st. millennium B.C. The French archaeological mission in the UAE that worked in the al Madam Plain from 1993 until 1995 undertook an extensive survey that identified c. 54 locations where archaeological remains were present. During the surveyFootnote 14 the French team located seven linear alignments of whitish mounds in sectors 2, 21, and 31. These linear alignments were identified as the remains of falaj access shafts that had been eroded over time. Interestingly, several of these alignments seemed to lead to the Iron Age settlement of al Madam 1-Thuqeibah.

Whilst we dug at the Iron Age village in sector 1 and studied its environment, objects, faunal and botanical remains, we never forget those alignments of whitish mounds, (located 2 km further east). The detection of the site’s main cultivation area near the outflow of the falaj, and archaeobotanical evidence for the cultivation of date palm at al-Madam during the Iron Age made it clear that an investigation of the falaj system was crucial to understanding water exploitation in this period (Fig. 4).

Fig. 4
figure 4

Aerial view of al Madam showing the archaeological sectors (Spanish archeological team, 2014)

Full size image

Al Madam 2 falaj: underground gallery

In the year 2002 we wanted to confirm, not only the existence of at least one falaj in the area, but also its relationship with the Iron Age village. The initial opening of a sounding around one of the whitish mounds located in sector 2, led to the finding of the first access shaft (in Arabic thuqba, pl. thuqāb). Since then, a 70 × 3 m trench allowed for the excavation of seven access shafts that led to the underground gallery. The mouths of these thuqāb had been directly dug into the sand and, after extending through a level formed by gravel, the wells entered the natural rock (Fig. 5). These mouths are protected by a sort of mortar made with material taken from the rock itselfFootnote 15 and have small steps carved into the walls, which would have allowed a person to climb down into the tunnel. We fully excavated thuqāb Tqb-1, 2 and 3, but Tqb 4 was dug in a section. The results indicate that the access shafts were blocked-up on purpose, probably after the drying up of the structure (Fig. 6). A progressive infilling of the tunnel, can be clearly seen in its stratigraphy, and is also suggested by the inner empty sections that are visible in the centre of the underground gallery. Indeed, the latter was partly infilled by two meters of loose wind-blown sand. At a certain point, which is difficult to date, the wells were, however, completely blocked, leaving the empty spaces preserved in the tunnel. This can be deduced from the features of the inner slopes.

Fig. 5
figure 5

Section of Thuqba 1 and falaj 1, found at al Madam 2. (M.A. Núñez Villanueva and I. Martín Gútiez, Spanish Archeological Team.)

Full size image
Fig. 6
figure 6

Section of Thuqbah 4 from surface until the beginning of the gallery. Al Madan 2, falaj 1. (M.A. Núñez Villanueva and I. Martín Gútiez, Spanish Archeological Team.)

Full size image

The vault of the falaj gallery is 0.70 m below the mouth of the access shaft. The vault is carved into the natural rock in the same way as the wells found in the village of AM 1, that is, with a very thin sharp pickaxe whose traces are visible in the rock. After first removing the sand between thuqāb Tqb 1 and Tqb 2 we found an angled tunnel with a width of 0.5–0.55 m and a total height of 4.80 m (Fig. 7). Eventually we cleared 35 m of gallery, the course of which follows a zigzag pattern. As such, its design differs from later falaj and from the traditional Iranian qanat that follow a straight-line course for several kilometres from the head of the qanat—the point where it meets the water table—to the outflow (Fig. 8). Several hypotheses can be put forward to explain the specific layout of al Madam falaj. Without an accurate measuring device, Iron Age diggers may have faced difficulties in estimating directions underground (Boucharlat 2017b). Alternatively, we suggest that the meandering trail of the gallery at al Madam would have permitted the collection of a greater amount of water than a straight tunnel. This would suggest that water was collected all along the whole water table not just at its headwaters, although it is unclear at the moment if the portion of gallery excavated corresponds to the infiltration or transportation section. The mother well of the falaj was not identified, as it is hidden by sand dunes near the border with the Sultanate of Oman; however, the survey suggests that it was located in a small depression situated to the southeast of the site, at the confluence of several wadis. Therefore, the water table exploited during the Iron Age came from wadi underflows or shallow aquifers (Boucharlat 2017b, p. 147). 

Fig. 7
figure 7

Underground gallery of al Madam falaj. (Spanish Archeological Team)

Full size image
Fig. 8
figure 8

Section and plan of the gallery of captage located in al Madam 2 after 2004 season. (M. Nuñez Villanueva, Spanish Archeological Team)

Full size image

The incredible height of the gallery suggests that water depletion due to increased aridity, overexploitation or a combination of both, was a concern, and that actions were undertaken to overcome it; this is also evidenced by the structures in the village.Footnote 16 Excavation also revealed that the gallery of the Iron Age falaj had been undoubtedly deepened. Initially, the bottom of the gallery was located 1.50 m below ground; it was later deepened to reach 4.80 m, in an attempt to find water (Fig. 9).

Fig. 9
figure 9

Underground gallery of the falaj showing the two stages of construction close Thuqba 2. (Spanish Archeological Team)

Full size image

As in the case of traditional aflāj, the amount of water that it extracts from the water table is never very abundant; Because the falaj would have specifically fed the gardens located outside of the village, another source would have been needed to supply water for people and cattle. Therefore, two wells were constructed within the village. We are certain however that the Iron Age village and falaj were contemporary.

Al Madam 2: Outflow of the falaj. Irrigation Chanel Network Area (ICNA)

In 2009, a geomagnetic survey identified the course of the falaj from the excavated thuqāb, to its downstream end. Sadly, the archaeological landscape of al Madam has been affected by a recently constructed road that fortunately has been elevated at the point where it crosses the falaj. Thanks to the intervention of the Department of Antiquities and Heritage of Sharjah Emirate, the road has not harmed the structure. North of this road a georadar survey, has also been undertaken that allowed us to identify the possible outflow of the falaj. Following the results of these geomagnetic and georadar surveys, seventeen soundings were dug along the course of the falaj in 2011–2014—which is 3 m deep at the level of the new road—up to the garden area.Footnote 17 Thanks to this work, it was possible to determine that when it emerges at the surface, the falaj became a deep open trench that ran for about 200 m before reaching the gardens. At this point, the falaj was about 20 cm deep.

From the primary channel, which had the same orientation as the gallery, a series of secondary channels branched off at right angles. These conducted the water to a number of ponds, pools and cultivated fields located on both sides of the main channel. This proved to be a genuine network of irrigation channels dug into the natural rockFootnote 18 (Fig. 10).

Fig. 10
figure 10

Aerial view of Ancient Irrigation Network area at al Madam 2 from northwest. (Spanish Archeological Team)

Full size image

During the season carried out in March of 2014, we located at least 32 secondary channels that connect hole-lines, which were interpreted as tree-pits; at the present time, we have uncovered 56 tree-pits. We also unearthed five large ponds, 2 m wide and approximately 30 m long, connected to the main channel through secondary channels. These ponds could be water reservoirs. The water was controlled by sluice gates made of local stone. We were also able to determine that the entire area had been modified and deepened at least twice just as the underground gallery of the falaj had. Due to a lack of water some adjacent tree-pits were also converted into ponds and the main channel was deepened. The reorganization of the network of channels was however never completed, as the water table became too deep and the area was abandoned.

As this area is still under excavation its overall dimensions are unknown to us, as is the storage capacity of the variously shaped ponds or the volume of water coming from channels with different widths and depths. However, we have observed that the cultivation area is framed by some raised mounds, still visible today, which might have resulted from the process of flattening the area for irrigation and dumping the removed soils at its edges. Taking into account this premise, we opened a transverse sounding between soundings 7 and 8 in February 2015, at the point where the eastern raised mound is located. Here, we were able to establish the eastern limit of the cultivated area as we unearthed the most upstream water distributor and secondary channels.

During the last season, topographic field studies allowed us to estimate that the Irrigation Channel Network Area (ICNA, that is, the field system located on both sides of the main channel) covered approximately 15 ha (600 × 300 m). However, up to now we have excavated only 20% of the ICNA (Fig. 11).

Fig. 11
figure 11

Sketch of falaj systems founded at al Madam 1 following the archaeological record. (M. Nuñez Villanueva, Spanish Archeological Team)

Full size image

This kind of irrigated garden area related to an Iron Age falaj is unique in the region and in the wider Oman Peninsula where groundwater draining systems are very well studied in the Late Islamic period, but for which many questions remain to be solved for earlier phases. Our example is a well-defined and articulated network, which is very similar to those from other arid regions such as southern Iraq, and from traditional oases of the United Arab Emirates. It is composed of lines of trees—usually date palm trees—protecting low crops with their shadow. In the present-day oases, date palms survive with little water and grow side by side with cereals, vegetables and fruits.

The modern inhabitants of al Madam have informed us that their ancestors grew crops directly upon the sand, suggesting that such a harvest was possible. While it was evident that the archaeological record would not be able to offer us evidence of crops grown on sand, amazingly, the system of water canalization and storage that irrigated these fields has remained preserved, excavated in the rock, since the beginning of the 1st millennium B.C.

Interestingly, the area also provided us with an abundance of Iron Age II potsherds that are clearly related to the material found at the nearby village of al Madam 1. Besides the pottery, considerable amounts of molluscs were also collected in all the channels.Footnote 19 They belonged to the family of the Thiaridae (Melanoides tuberculata and Truncatella marginata), suggesting a constant presence of clean and fresh flowing water in the whole network. At the outflow of the falaj, besides the Thiaridae, we also found two date stones and the remains of Terebralia palustris, which, after calibration with the marine calibration curve “marin09.14c” permitted us to date the structure to some point between 1065 BC and 808 BC.Footnote 20

Conclusion

The population of al Madam chose to remain in a very hostile environment that nevertheless could be generous if water was well managed. The evidence indicates that the Iron Age people of al Madam conducted groundwater to their fields through tunnels, channels, ponds and pits. It is evident that this community had an advanced knowledge of the subsoil: community members knew which natural rock could be dug into with minimal difficulty and where the water table was most easily reached. In fact, water probably oozed out of some of the depressions we find at the head of the falaj. This would have provided a clue to the water’s depth below the surface. Al Madam’s inhabitants knew how to maximise the potential of their environment, combining agriculture, cattle breeding, trade and hunting to remain in a place that, once abandoned, would never be inhabited again due to a decrease in the water table that clearly occurred sometime in the Iron Age. The hydraulic knowledge of the Iron Age inhabitants of this settlement continues to amaze us, and further reveals itself with every excavation campaign we undertake at al Madam.