Abdullakhan Bandi: The 1580s Stone Dam Near Nurota
How a 1580s Bukhara khan's stone dam in the Beklarsoy gorge stored about 1.2 million m3 of floodwater: shaft sluice, lime-and-ash mortar, UNESCO listing.
Fifteen metres of cut stone, 85 metres long at the crest, laid across a gorge on the south side of the Nuratau range in the 1580s. That is Abdullakhan Bandi, the dam that Abdullah Khan II of Bukhara ordered to catch the spring floods of the Beklarsoy stream. It has sat on Uzbekistan's UNESCO Tentative List since 18 January 2008, yet very little about it is available in English beyond that listing and one short journal paper.
This post is built from those documents only: the UNESCO entry (ref. 5288), a 2023 paper by Uralov Akbarali and Bakhodirov Abrorkhan of the Samarkand State University of Architecture and Construction, the separate UNESCO entry for the older Khanbandi dam, and a 2024 paper on Khanbandi. We haven't been to the site, we can't tell you how to reach it, and the sources disagree on several numbers. Where they do, you'll see both figures.
Where it is, and why the UNESCO entry gets that wrong
The 2023 paper puts the surviving remains in the Beklarsoy gorge near the village of Eski Okchob, in Koshrabot district of Samarkand region, about 65 km east of Nurota. The UNESCO entry, which titles the property "Abdulkhan Bandi Dam", agrees on the stream (it spells it Biglarsay) and says it flows down from the southern side of the Nurata range.
Then the same entry lists the property under "Jizzakh Region, Forish District", with coordinates written as "N66 42 E40 29". Two things are off. The figures are reversed (66°42' is a longitude and 40°29' a latitude, not the other way round), and they're identical to the coordinates on the Khanbandi entry, a different dam in Forish district that Uzbekistan listed in 1996. Our reading is that the 2008 entry copied Khanbandi's location block. Don't plug those numbers into a map expecting to find Abdullakhan Bandi.
For context on the wider area, the town at the western end of the range is covered in our Nurata page on the spring, fortress and petroglyphs, and the ridge itself in the Nuratau mountains guide.
Why the khan built it: the Abdullanoma account
The story comes from the Abdullanoma, a Persian chronicle of 16th-century Mawarannahr written between 1584 and 1590 by the Bukhara historian Hafiz Tanish Mirmuhammad al-Bukhari. The 2023 paper cites a manuscript held by the Uzbek Academy of Sciences' Institute of Oriental Studies (inventory no. 2207).
As the paper summarises it, the khan was on his way home from a military campaign when he came through Okchob, close to Josh village. Floodwater gathered there every spring and was hardly used for farming. He ordered a large dam across the Okchob valley, the gorge now called Beklarsoy, so the floods could be stored and released onto fields. (The English translation names the campaign's destination as "Istan"; we couldn't pin down what that refers to, so we've left it out.)
Who actually ran the job? The UNESCO entry says, citing a historian it calls "Khafizi", that the dam was built by Akhmadali Nayman atalik on Abdullah Khan's order. Atalik was a high court rank in the Bukhara khanate. That attribution rests on the UNESCO text; the 2023 paper doesn't name a builder.
The dam fits a pattern. The 2023 paper credits his reign with many canals as well, naming the Tuyatortar channel among them, and his name still hangs on other public works, such as the domed silk hall described in our walk through Bukhara's trading domes, and the Shaybanid building boom also produced Tashkent's Kukeldash madrasah around 1569 to 1570.
How the dam worked
Both sources describe a full water-control complex rather than a simple wall. UNESCO lists four parts: the dam, a floodgate worked with beams, an underground gallery with a vertical shaft, and an emergency ("catastrophic") spillway.
The wall
The upstream face, the one holding back the water, was built almost vertical. The downstream side steps back like a staircase. The 2023 paper explains the logic: the stepped back resists the pushing and overturning force of a reservoir about 15 metres deep, which is why the base is roughly 3.5 times thicker than the top. UNESCO says the stone was worked flagstone; the paper says shale.
The shaft sluice and two-tier outlet
This is the clever bit. A vertical shaft, 14 m deep and 2 m across, sat in the centre of the dam. Two outlets fed into it. The lower one was a horizontal tunnel with a pointed (lancet) arch, set 1 m above the base, 19 m long, 50 to 60 cm wide and 1 m high. The upper one started 1 m higher: a narrow, deep slot about 50 cm wide and 12.5 tall, also connected to the shaft. The entry drops the unit there; metres is our reading, since a slot starting about 2 m up and rising 12.5 m tops out at the dam's 14.5 to 15 m.
Halfway along that slot, the stone has cut slots for wooden beams. Stack beams in and the opening shrinks; pull them out and more water drops into the shaft and out through the tunnel. In engineering terms that works like a stop-log gate, a principle still used on small irrigation structures, and it let the keepers release water by level rather than all at once.
The emergency spillway
On the right side of the dam, separate from the outlets, UNESCO records a spillway 1 m wide and 1 m deep to throw off excess water in a flood. The 2023 paper describes the same feature as a structure that "ejected excess water to the outside".
The numbers, and where they disagree
Length is the easy part: both sources give 73 m at the bottom and 85 m at the top. Base thickness is 15.3 m in both. After that, they drift.
- Height: UNESCO says 15 m. The 2023 paper says 14.5 m above the ground, then, a sentence later, 15 m high. Read it as 14.5 to 15 m.
- Crest thickness: UNESCO says 4.6 m; the paper says 4.5 m.
- Storage: the paper puts the water held behind the dam at roughly 1.2 million cubic metres. UNESCO gives no figure.
- Efficiency: the paper adds that the reservoir caught only about 10% of the stream's annual flow.
According to the 2023 paper, the water served seven villages in what is now Koshrabot district: Kamar, Okchob, Urganji, Josh, Ravot, Jilontamgali and Soykechar, and brought 1,000 to 1,200 hectares under cultivation. The authors note that people there still grow grapes and other crops.
The mortar: lime, plant ash and a camel-milk story
With no cement available, the question both sources circle is what held the stones together underwater. UNESCO just says "water-resistance mortar". The 2023 paper's text calls it a special water-resistant ganch mixture (ganch is the local gypsum plaster), then repeats a story from "historical sources": masons cut stone in the mountains, builders prepared alabaster, and the alabaster powder was boiled in molasses with camel milk brought in wineskins.
The paper's own lab table tells a different story. Citing chemical analyses of samples by N.S. Grajdankina, it lists Abdullakhan Bandi's binder as 55% lime and 45% plant ash, with no gypsum at all. (The English version labels the lime column "bleach", a likely mistranslation.) The paper's summary sentence after the table, about lime mixed with sand or vegetable oil, doesn't match the table either. Our take: trust the table, and treat the camel milk as tradition.
| Dam | Century | Location | Height | Binder (2023 paper's table) | Water held |
|---|---|---|---|---|---|
| Khanbandi | 10th | Forish district, Jizzakh | 15.25 m | 50% lime, 50% earth | about 1.6 million m3 |
| Gishtband | 12th | Not given in the 2023 paper | Not given | 80% lime, 20% plant ash | Not given |
| Abdullakhan Bandi | 16th (1580s) | Koshrabot district, Samarkand | 14.5 to 15 m | 55% lime, 45% plant ash | about 1.2 million m3 |
Sources for the table: Khanbandi's height comes from its UNESCO entry, which also describes nine cone-shaped outlets at different heights and a base four times thicker than the top. Its 1.6 million m3 figure is from the 2024 paper by Axmadkulov Nozimjon. The binder columns are from the 2023 paper. Gishtband is the weakest row; the 2023 paper gives only its century and mortar.
The Pascal claim, and what we make of it
Uzbek papers on these dams repeat a claim that the 10th-century builders of Khanbandi used the principle of fluid pressure that Blaise Pascal described seven centuries later. The 2023 paper says so in its introduction, and the 2024 paper makes it the theme of its abstract. The claim is about Khanbandi, not Abdullakhan Bandi, and it's the authors' interpretation: tiered outlets show the builders understood that pressure rises with depth, which is a long way from a stated physical law. We're reporting it, not endorsing it.
Research history and listing status
The record is short:
- 1913: remains first recorded (UNESCO).
- 1957 to 1962: studied by the Makhondarya archaeological team of the Uzbek Academy of Sciences' Institute of History and Archaeology (UNESCO names 1957 and 1962).
- 18 January 2008: submitted to the Tentative List by Uzbekistan's National Commission for UNESCO, under criterion (iv): an outstanding example of a type of building or technological ensemble that illustrates a significant stage in human history.
- August 2022: field survey of the surviving section by the Samarkand team behind the 2023 paper.
Two clarifications. The 2023 paper says the dam was "included in the UNESCO World Heritage List" in 2008. It wasn't; a Tentative List entry is only a country's notice that it may nominate a site later. And the old dam no longer stores water. The paper says it was replaced by the modern Okchob reservoir, and UNESCO notes the monument has only been protected since that new dam was built.
This isn't a visiting guide. Nothing we read confirms public access, signage or a road to the remains, so if you want to see it, ask a guide in Nurota or Samarkand to check locally first.
Primary documents: the UNESCO Tentative List entry for the dam, the Khanbandi entry it borrows coordinates from, the 2023 IJTSRD paper by Uralov and Bakhodirov (PDF), and the 2024 Khanbandi paper (archived copy; the conference site has since gone offline). All checked October 2026.