Aral Sea Restoration: Can It Come Back? The Numbers
Aral Sea restoration by the numbers: why Kazakhstan's North Aral recovered after the Kokaral dam, why the south can't, and what Uzbek saxaul planting does.
No. The southern Aral Sea, the part that belonged to Uzbekistan's fishing towns, isn't coming back in any lifetime that matters to someone reading this. The small northern lobe in Kazakhstan is a different story: it has already partly returned, and a second round of work planned for 2026 to 2029 aims to push it further. Aral Sea restoration is really two questions with two answers, and most headlines blur them.
Uzbekistan, meanwhile, isn't refilling anything. It's planting shrubs across the old seabed on a colossal scale, and the official numbers and the independent ones don't quite agree.
Two seas, two scoreboards
In 1960 the Aral covered about 68,000 square kilometres, and its water held around 10 grams of salt per litre, less than a third as salty as the ocean, according to the US Geological Survey. By the late 1980s it had split in two. The North (Small) Aral, fed by the Syr Darya, is in Kazakhstan. The South (Large) Aral depends on the Amu Darya and straddles the Kazakh-Uzbek border, its Uzbek shore in Karakalpakstan.
Here's where each stands on the figures we could source (last checked: September 2026).
| Measure | North Aral (Kazakhstan) | South Aral (Kazakhstan/Uzbekistan) |
|---|---|---|
| Main river | Syr Darya | Amu Darya (reaches it only intermittently) |
| Salinity | About 8 g/L after the dam (White & Micklin, 2021); Kazakh officials cited 23 g/L in 2025 | Above 100 g/L, roughly three times seawater (USGS) |
| Water volume | 27 km³ in 2025, up 42% (Kazakh ministry) | No agreed current figure; eastern basin went fully dry in 2014 |
| Surface area | 2,804 km² before the dam, ~3,300 km² after | A fraction of its 1960 extent; varies year to year |
| Fish catch | 695 t in 2005, ~6,000 t by 2016, ~7,800 t in 2024, ~3,900 t in 2025 | No commercial fishery |
| Next step | Raise the level from 40.4 m to 44 m, 2026–2029 | No plan to refill; seabed planting instead |
The North Aral salinity gap in that table is real, not a typo. The lower figure is a peer-reviewed average from just after the dam closed; the higher one is what Kazakhstan's water ministry quoted in May 2025, alongside a 10 g/L target. Treat 8–23 g/L as the honest range.
What the Kokaral dam actually did
It's a modest structure for its reputation. The Kokaral (Kok-Aral) dike runs 12–13 km (sources differ) across the old strait between the two seas, with a concrete control section and nine spillways that let surplus water spill south when the Syr Darya runs high. It was finished in late summer 2005 under the World Bank's Syr Darya Control and Northern Aral Sea Phase 1 project, which cost roughly $86 million, with about $64.5 million financed by the Bank, according to geographers Kristopher White and Philip Micklin in a 2021 review of the northern recovery.
The sea hit its 42-metre design level in March 2006, six months after the gates closed. Experts had expected it to take years.
That speed explains why the north worked. The dam added no water; it stopped the Syr Darya draining south into a shallow basin to evaporate, and kept it in a smaller, deeper bowl. A 2-metre rise in level (about 4.5%) grew the surface by around 18%. Salinity fell to something close to 1960 conditions, and 18 fish species came back, including pike-perch, bream and carp. The harvest went from 695 tonnes in 2005 to about 6,000 tonnes by 2016, per the same review.
Our take: it's one of the best-value environmental projects of the century so far. It's also a trade. By sealing the north, Kazakhstan effectively cut off most of the Syr Darya water that used to leak south, which is one reason the Large Aral's eastern basin dried out completely in 2014.
Why the north's comeback is only half finished
The dam was always a compromise height. Holding the level at 42 m saved the sea but didn't bring the water back to Aralsk, the old Kazakh port, which is still well short of its former shoreline. And the level hasn't stayed put: the Phase 2 plan announced with the World Bank talks about raising it from 40.4 m to 44 m on the Baltic height system, which suggests it has slipped below the original design mark.
Phase 2, scheduled for 2026 to 2029, would rebuild and heighten Kokaral, expand the surface to 3,913 km² and lift the volume from 27 to 34 km³, officials told reporters in May 2025. A feasibility study was due by December 2025, and in January 2026 the Astana Times reported the project was "undergoing national expert review", without publishing a final price.
The catch: a higher dam only helps if water arrives. Officials said 2.6 billion m³ reached the sea in 2024, yet Global Voices reported in May 2026 that the 2025 fish catch halved, from about 7,800 to 3,900 tonnes. Even good years land near a tenth of the 80,000 tonnes the whole Aral produced before 1960.
Why nobody is trying the same trick in the south
Could Uzbekistan just build its own dam? It's the obvious question, and the answer is mostly about evaporation.
The USGS puts rainfall over the Aral at under 200 mm a year against 1,000–1,200 mm of evaporation. A sea in that climate lives on river inflow, and the Amu Darya is almost fully spoken for by irrigation upstream long before it reaches Karakalpakstan. White and Micklin note it only "periodically" reaches the southern sea at all. A dam holds water. It can't conjure it.
Geography doesn't help. The eastern half of the Large Aral was always a wide, shallow plate that loses water fast; the deep western basin under the Ustyurt cliffs is what's left. Refilling toward the 1960 shore would mean sending tens of cubic kilometres a year downstream instead of onto the cotton and wheat fields millions of people depend on. No government in the basin has proposed it.
Then there's the new pressure upstream. Afghanistan's Qosh Tepa canal, planned at 285 km and roughly half built by September 2025, is expected to take as much as 10 km³ a year from the Amu Darya once it's fully running around 2028, according to the Carnegie Endowment. Estimates of its share of the river range from a fifth to a third. Any of those moves the south further from recovery.
The forest on the seabed: two million hectares, or fewer?
Since Uzbekistan can't bring the water back, it's trying to pin the ground down. The exposed bed, now the Aralkum desert, throws salt and chemical-laced dust across Karakalpakstan. The government's answer, from around 2018, was mass planting of black saxaul (Haloxylon), a drought-hardy shrub-tree with deep roots, alongside tamarisk and other salt-tolerant species, often sown from aircraft.
The official numbers have climbed fast. Mongabay reported in May 2024 that 1.7 million hectares had been planted in five years, with a 2024 target of 150,000–200,000 more. By July 2026, Gazeta.uz reported "over 2 million hectares" of new plantations on the dried bed, at a meeting where President Mirziyoyev also set a 2026–2030 goal of creating or restoring 1.27 million hectares of forest nationally (last checked: September 2026).
Now the independent side.
- Hectares treated isn't hectares of forest. The two-million figure counts area sown or planted, not how many plants took or how dense the cover is. These are dust barriers of scattered shrubs, not woodland.
- Survival swings with rain. Turaxon Ziyotov of Tashkent's Forest Research Institute told Mongabay that survival runs 50–70% in wetter years and drops to around 30% in dry ones.
- Across the border, the measured range is enormous. A 2021 field study in the Central Asian Journal of Water Research examined 24 saxaul plantations planted in 2009–2019 on the Kazakh side and found survival anywhere from 0.12% to 78%. Sandy loam did well; crusted takyr and salty clay did badly.
- The carefully counted plots are small. UNDP's crowdfunded "Green Aral Sea" project, which counts individual seedlings, reported 823,000 trees on 658.2 hectares over five years, and 80,000 seedlings on 80 ha near Moynaq in March 2025.
The fair reading: the planting is real, huge in area, and some of it clearly works. How much of the two million hectares is living, dust-stopping cover hasn't been independently published, and forestry specialist Dildora Aralova warned Mongabay that a serious drought could undo much of it.
What "restoration" means on the Uzbek side
Because the sea itself is off the table, recovery in Karakalpakstan is mostly measured in people. The UN's Multi-Partner Human Security Trust Fund for the Aral Sea Region was launched on 27 November 2018; Uzbekistan pledged $6.5 million to it, with Norway, Finland, the EU and South Korea among other donors. Its first call approved two projects worth $3.085 million, and the fund says more than 293,000 residents had better access to health, food and economic security by 2024.
Modest sums next to one $86 million Kazakh dam, and they buy different things: water pipes, clinics, small farm grants. The July 2026 "desert economy" plan adds halophyte crops, desert nurseries, pistachio plantations and ecotourism. Whether that becomes income in Moynaq or stays a ministry slide is the real test.
For the human side rather than the hydrology, start with our piece on Karakalpak people and their culture, and use the wider Karakalpakstan travel guide to plan the republic as a whole.
What a visitor can actually see of all this
Almost none of the good news is visible from Uzbekistan. The recovering North Aral is only reachable from the Kazakh side, around Aralsk and the Kokaral dike itself, which means a border crossing and a very long drive from Nukus; there's no practical way to see both seas on one Uzbekistan trip.
On the Uzbek side you see the other half of the story. Seabed tours out of Moynaq drive past the saxaul plantations, so you can judge the cover for yourself. The old harbour at Moynaq, covered in our guide to the ship graveyard, shows where the shoreline once was. The only surviving southern water, the salty western basin, needs a multi-day Ustyurt Plateau expedition to reach.
Frequently Asked Questions
Is the Aral Sea growing again?
Only the northern part. Kazakhstan's North Aral grew after the Kokaral dam closed in 2005, reaching about 3,300 km² within a few years and holding 27 km³ of water by 2025, and a 2026–2029 upgrade aims for 3,913 km². The southern Aral, shared with Uzbekistan, has kept shrinking and its eastern basin dried out completely in 2014.
Why can't Uzbekistan build a dam to save the South Aral?
Because there's no water to hold. Evaporation over the Aral runs 1,000–1,200 mm a year against under 200 mm of rain, and the Amu Darya is almost entirely used for irrigation before it arrives. A dam only keeps inflow from spreading out and evaporating; the south barely gets any inflow to keep.
How much of the dried Aral seabed has Uzbekistan planted?
Officially, over 2 million hectares by 2026, up from 1.7 million reported in May 2024. That's area planted, not proven forest. Researchers put saxaul survival at roughly 30–70% depending on rainfall, and a Kazakh field study found 0.12–78% across different sites, so the living cover is certainly smaller than the headline.
Will the Aral Sea ever fully recover?
Not to its 1960 size of about 68,000 km². That would need tens of cubic kilometres of river water a year taken away from farming in several countries, and Afghanistan's Qosh Tepa canal will draw more from the Amu Darya, not less. A healthier North Aral and a planted, stabilised southern seabed are the realistic ceiling.