Avdat National Park

The Nabataeans reached this ridge as caravanners rather than settlers. English Wikipedia, drawing on the excavation literature, describes Avdat in the 3rd to late 2nd century BCE as a seasonal camping ground on the early Petra to Gaza road, and the parks authority's own brochure numbers it station 62 on a route it traces some 2,400 kilometres from Oman and Yemen through Saudi Arabia and Jordan to the Mediterranean.

The town takes its name from a deified king. Avraham Negev, who excavated here, quotes the 6th-century grammarian Stephanus of Byzantium: 'Oboda, a place of Nabateans. Uranios, in his fourth book on the Arabs, writes where king Obodas, who is deified, is buried.' Negev argued that Obodas I, who ruled around 96 to 85 BCE, built the new city, named it after himself and was worshipped there after his death. The Encyclopaedia Judaica entry is more cautious and says only that the site was probably named after the deified king Obodas.

Its wealth peaked with the kingdom's. Under Aretas IV, from 9 BCE to 40 CE, the incense trade enriched the Nabataeans and their main centres developed: Petra in Transjordan, Avdat and Halutza in the Negev. The parks authority sets out what a caravan of that road actually was, which is more prosaic than the word suggests: ten to twenty camels carrying 150 to 200 kilos each, every animal led by a handler on foot, moving at a pace set by the spacing between them.

Then the trade left. Rome annexed the kingdom in 106 CE, and by the early to mid 3rd century international traffic through the region had ceased. Avdat did what the other Negev highland towns did and turned to farming, and above all to wine. The Nabataeans, whose earliest agricultural inscriptions at this site date to 88 to 98 CE, were still here: recent excavation of a residential quarter outside the town wall found Nabataean pottery in production into the middle of the 3rd century.

• Seasonal Nabataean camping ground, 3rd to late 2nd century BCE
• Station 62 on a route of about 2,400 km, Oman and Yemen to Gaza
• Height of the trade under Aretas IV, 9 BCE to 40 CE
• Roman annexation in 106 CE; international trade through the Negev ends in the 3rd century
• The town survives by farming and winemaking rather than by transit

Everything that made Avdat important sits on a platform along the western edge of the ridge, laid out before the end of the 1st century BCE. The Nabataean temple precinct is at its western end; the Temple of Oboda, dedicated to the deified king, is a tripartite building of porch, hall and adytum measuring about 14 by 11 metres. An inscription found nearby names Aphrodite, the Greek equivalent of the Nabataean goddess Uzza, and excavation of the temple walls turned up the moulded stucco and frescos that Nabataean temples usually carried.

The dated stones are the most specific thing here. The temple gate lintel bears a Nabataean dedication of 268 CE to Zeus-Oboda, naming its dedicators and 'all friends of Oboda'. The lookout tower above the upper car park carries a Greek inscription asking Zeus Oboda to help Irenius, who built it with the architect Wailos of Petra, in a year counted from the founding of Provincia Arabia in 106 CE: the parks authority reads that as 294 CE, and Wikipedia as 293/294.

The Christian town reused the same ground. In the 5th and 6th centuries a citadel and a monastery with two churches were built on the acropolis. The southern church, part of the monastery, is named for Saint Theodoros after the earliest of the inscribed marble slabs set into its floor, which run from 541 to 618 CE. The northern church is the earlier of the two and held the town's cathedral, and behind it stands a cruciform baptismal basin built into the remains of a Nabataean temple.

Two details in the baptistery are worth stopping for. The account published through the Jewish Virtual Library describes two fonts: a large cross-shaped basin for adults and a smaller round one, which it reads as being for babies. Standing at the temple gate immediately outside, a visitor is looking at a Nabataean lintel of 268 CE, a Christian font a few metres behind it, and the valley the town's vineyards stood in, all within one turn of the head.

• Temple platform laid out before the end of the 1st century BCE
• Temple of Oboda, about 14 by 11 m, on the acropolis
• Temple gate lintel dedicated to Zeus-Oboda, 268 CE
• Roman lookout tower inscription read as 294 CE
• Two churches and a citadel added in the 5th and 6th centuries
• Floor inscriptions in the southern church dated 541 to 618 CE

The excavation history is short and well documented, and the parks authority prints it on the site brochure. Michael Avi-Yonah uncovered Avdat in 1958 and Avraham Negev worked here from 1959 to 1961, both for the Hebrew University of Jerusalem. Conservation and reconstruction were carried out in the 1960s. Negev returned with Rudolph Cohen from 1975 to 1977 and alone in 1989; Ofer Katz, Gil Tahal, Tali Erickson-Gini and Peter Fabian dug for the Israel Antiquities Authority from 1990 to 1994, and Erickson-Gini and Fabian again in 1999 and 2000.

So a good deal of what stands has been stood back up. The marble baptistery behind the northern church is described by the parks authority as reconstructed. The winepress on the short route is a reconstruction. The Roman villa on the drive up the hill has been rebuilt around its courtyard cistern. None of this is concealed: the brochure labels the reconstructed elements, and the visitor centre holds a scale model of the Byzantine city precisely so that the ruin does not have to be read as if it were complete.

Then there is the damage nobody planned for. On the parks authority's account, vandals destroyed remnants of the Byzantine settlement on 5 October 2009, toppling, smashing and spray-painting buildings, vaults and columns. Ynet reported that hundreds of artefacts were smashed, that walls, a staircase and an ancient winepress were painted over, and quoted the authority's chairman, Eli Amitay, saying that a World Heritage Site had received a grave blow. Two cousins from the Azazma were later charged, with the indictment reported as putting the damage at 8.7 million shekels.

The repair took three years. The parks authority ran the project with the Israel Antiquities Authority and government funding: ruins restored, archaeological remains conserved, new signage and interpretation along the visitor route, and a security system installed against a repeat. The arches of the vaulted room in the Byzantine quarter, which once bore stone roof slabs, were re-erected during that project after the vandals pulled them down. Much of the loss was not recoverable, and the site does not pretend otherwise.

• Uncovered by Michael Avi-Yonah, 1958; excavated by Avraham Negev, 1959 to 1961
• Conservation and reconstruction carried out in the 1960s
• Israel Antiquities Authority seasons in 1990 to 1994 and 1999 to 2000
• 5 October 2009: vandals wreck Byzantine remains, columns and a winepress
• Three-year restoration follows, with new signage and a security system

The version on the signs is the simplest. In the early 7th century a huge earthquake struck the area, destroyed Avdat and led to its abandonment. The evidence is visible on the walk: the arches in the Byzantine quarter lean, and the brochure points at them as the physical record of that shock. Korjenkov and Mazor's archaeoseismic surveys, published in 1999, separated 7th-century seismic effects at Avdat from those of an earlier earthquake.

The Encyclopaedia Judaica tells it differently. In that account the two churches were razed during the Persian invasion of 614 CE, and the city was completely deserted after the middle of the 7th century. The seismic sequence itself has also been rewritten more than once. Erickson-Gini set out four events, in the early 2nd century, in 363, in the early 5th century and in the early 7th; the 'Atiqot report on the residential quarter argues that 363 did only minor damage here and that the serious destruction came a few decades later, in the early 5th century, on a date nobody can fix.

The most recent work removes the ending altogether. The Avdat in Late Antiquity Project, reported in Antiquity in 2022 by Scott Bucking, Daniel Fuks, Zachary Dunseth, Lior Schwimer and Tali Erickson-Gini, investigated a two-hectare residential sector of rock-hewn dwellings on the southern slope. Nine radiocarbon samples from dung-enriched deposits came back Umayyad to Abbasid, 650 to 890 CE, corroborated by pottery from the Khirbet al-Mafjar kilns. The Dead Sea earthquake catalogue's page on the site now allows abandonment as late as the 9th or 10th century.

This is not a quibble about a century. It changes what the caves on the slope were: not the wreckage a departing population left, but stabling and storage still being used two hundred years after the town is supposed to have died. Researchers still differ over how much weight to give the earthquake, and the site's own interpretation has not caught up with the radiocarbon.

• Parks authority: destroyed by an early 7th-century earthquake and abandoned
• Encyclopaedia Judaica: churches razed in the Persian invasion of 614 CE, deserted after the mid-7th century
• Erickson-Gini: four earthquakes, with the serious damage in the early 5th century rather than in 363
• Antiquity, 2022: radiocarbon dates of 650 to 890 CE from the cave quarter
• Leaning arches in the Byzantine quarter are the visible evidence

The reason Avdat survived the collapse of the caravan trade is in the valley below it. At least five Byzantine winepresses have been found at the site, and the parks authority's Hebrew page now routes visitors along a winepress trail past them, including a sophisticated industrial press on the south side with three collecting vats holding some 6,000 litres. The earliest press at or near Avdat belonged to a farmhouse of the late Roman or early Byzantine period, and the grapes were treaded where they grew, at the foot of the town.

Water was engineered rather than found. Rainwater ran off the hill through channels into large cisterns, one of them fed by a channel at the foot of the citadel's southeastern tower, and surface runoff was caught in more cisterns across the surrounding ground. The Byzantine bathhouse below the site, built in the 4th century by Roman soldiers, drew on a well tunnelled through bedrock beside it. Its depth is given as 70 metres in the parks authority's brochure and as about 64 metres on the authority's own web pages, which is a discrepancy the operator has not resolved.

The farming system was put back into service in living memory. In the late 1950s the Desert Runoff Farms Unit, now shared between the Hebrew University of Jerusalem and Ben-Gurion University of the Negev, reconstructed two Nabataean farms, one at Shivta and one at Avdat, under the ecologist Michael Evenari. The Avdat farm ran to about 20 acres and was planted with 3,000 orchard trees along with cereals, vegetables, legumes and pasture plants, watered by nothing but runoff steered off the slopes.

The vines came back too. In 2003 the parks authority planted a heritage vineyard beside the southern winepress, among the ancient terraces, reconstructing old grape varieties and Byzantine desert farming methods from the period when the Negev was known for its wine. It sits within a much larger picture: a survey of more than 500 sites across some 150 square kilometres around Avdat, published in 2024, treats the town and its agricultural hinterland as a single system rather than as a ruin with fields near it.

• At least five Byzantine winepresses, linked by a marked winepress trail
• Southern industrial press: three collecting vats of about 6,000 litres
• Bathhouse well given as 70 m in the brochure and about 64 m on the park's web pages
• Evenari's reconstructed runoff farm: about 20 acres, 3,000 orchard trees, from the late 1950s
• Heritage vineyard planted in 2003 between ancient terraces