Afghanistan Constructive Voices · Data Report

Transforming Landscapes: The Qush Tepa Canal and Northern Afghanistan's Ecosystems — the 2026 report

A 285-kilometre channel is being cut across a desert that already produces the region's dust storms, fed by a river already in decline. The engineering question and the ecological question have separate answers — and only one of them is being tracked.

Published August 2026 · Updated from March 2025 · Sources: FAO, OCHA, UNICEF/WHO, NDC Afghanistan, SIWI, Carnegie, Times of Central Asia, Asia-Plus/Azattyq, University of South Dakota

The canal is no longer a proposal. Excavation on the main channel is all but finished, the river it draws from is running at two-thirds of its normal flow, and there is still no agreement covering how the water is shared. This is what the 2026 data says about the Qush Tepa Canal and the landscapes it is about to reorganise.

This report updates our March 2025 analysis, originally written by Abdullah Hakimi for Afghan Architecture & Urbanism, and sits within our ongoing Afghanistan built environment coverage.

Rural landscape along the Balkh River, northern Afghanistan — Rural landscape along the Balkh River — the fertile margin between mountain-fed irrigation and the sand dunes to the north.

Northern Afghanistan is a land of contrasts. Vast deserts and fertile deltas, mountain-fed rivers and scattered wetlands sit within a few dozen kilometres of one another, and the boundaries between them have never been stable. What is new is the speed at which those boundaries are now being redrawn by a single piece of infrastructure.

When we published the original landscape assessment in March 2025, the Qush Tepa Canal was a project with 81 per cent of its second phase excavated and a completion date that observers treated as aspirational. Eighteen months on, the second phase is effectively dug, the intake works are largely built, and the argument has moved from whether the canal will happen to what happens when the water arrives.

The 2026 picture in seven numbers

285–290 kmLength of the main channel, from Kaldar in Balkh to Andkhoi in Faryab

98%Phase two excavation reported complete by Kabul, March 2026

66.8%Amu Darya flow as a share of normal, 11 February 2026 — against 101.8% a year earlier

550,000 haDesert to be converted to farmland — roughly 1.36 million acres

6–10 km³Water Afghanistan plans to divert annually once the system runs at capacity

2028Scheduled completion of the full system, including distribution canals

ZeroTreaties allocating Afghanistan a share of the Amu Darya

what is being built · what is missing or falling · fixed scale

01 / The build

Where construction actually stands

The canal is being delivered in three phases by Afghanistan's National Development Corporation, with the work carried out by dozens of contracted companies and, at peak, thousands of pieces of heavy plant. Phase one — the stretch from the Amu Darya intake at Kaldar to Dawlat Abad — was finished in 2023. Phase two, the long run west to Andkhoi, has absorbed most of the effort since.

Qush Tepa Canal: reported construction status by component
Component	Route / scope	Length	Reported status
Phase 1	Kaldar (Balkh) to Dawlat Abad	108 km	Complete, 2023
Phase 2	Dawlat Abad to Andkhoi (Faryab)	177 km	~95% excavated Oct 2025; 98% claimed Mar 2026
Intake / dam	Left bank of the Amu Darya, Kaldar	—	~85% Oct 2025
Crossings	Large and medium bridges along the route	—	Under construction through 2025–26
Phase 3	Secondary and distribution canals	500+ km	Not yet delivered at scale
Full system	Main channel plus distribution network	—	Target completion 2028

Sources: National Development Corporation site updates (October 2025); Islamic Emirate statements reported by TOLOnews and Ariana News (2025); progress claim of 98 per cent reported March 2026. Figures are as announced by the implementing authority and are not independently audited.

Independent monitoring broadly corroborates the pace, if not always the percentages. A satellite timeline study published by the University of South Dakota found roughly 70 per cent of excavation complete by the end of 2024 and the channel extending to some 230 kilometres by March 2025 — fast enough, the author noted, to make operation possible around 2026 rather than the official 2028. Afghan officials told delegates at Tashkent Water Week in March 2026 that more than 200 kilometres had been excavated.

The project's origins predate the current government by half a century. A canal on this alignment was first proposed under Mohammad Daoud Khan in the 1970s. The Ghani administration revived it, and in December 2018 signed a USAID-funded contract with the engineering firm AECOM for a feasibility study; by August 2021 the intake location had been identified and the documentation was largely complete. Construction began in March 2022. Cost estimates for the full scheme range from roughly $680 million to $2 billion, financed domestically.

02 / The river

A declining flow, and no agreement over it

The Amu Darya, with the Panj, runs some 2,400 kilometres and carries an annual runoff estimated at around 75 cubic kilometres. Roughly 30 per cent of that water originates on Afghan territory — about 21 to 23 cubic kilometres a year on a thirty-year average — yet Afghanistan has historically withdrawn only 3 to 5 cubic kilometres. Qush Tepa is designed to push that figure to somewhere between 9 and 11.

Afghanistan's legal position is straightforward and, in the abstract, difficult to argue with: it is a riparian state on a transboundary river and has never been party to the allocation framework that governs the rest of the basin.

Amu Darya allocation under the post-Soviet framework carried into the 1992 Almaty Agreement
State	Allocated share	Status
Uzbekistan	48.2%	Signatory
Turkmenistan	35.8%	Signatory
Tajikistan	15.4%	Signatory
Kyrgyzstan	0.6%	Signatory
Afghanistan	None	Not a party

Source: Azattyq Asia / Asia-Plus review of the Amu Darya allocation regime, August 2026. The Soviet-era quota system was preserved by the newly independent republics after 1991; Afghanistan was excluded from both the original protocol and its successor.

The arithmetic is being tested under conditions nobody planned for. As of 11 February 2026, flow in the Amu Darya stood at 66.8 per cent of its normal level, against 101.8 per cent at the same point a year earlier. The FAO reported that snowfall at the start of the 2025/26 wet season was at a 25-year low — a critical signal in a basin where crop and livestock production depend on snowmelt. Over the past three decades, the river's flow has fallen by roughly 12 per cent overall.

Against that backdrop, the diplomacy has been notably restrained. Uzbekistan and Afghanistan signed an agreement on joint management of Amu Darya basin water resources in spring 2025. Tashkent has repeatedly offered technical assistance, including — according to reporting by Rivers without Boundaries in March 2026 — help with the design and concreting of the canal bed. Uzbekistan's water resources minister has publicly accepted that Afghanistan has a right to the river. Turkmenistan, which depends on the Amu Darya more completely than any other state in the basin, did not send a delegation to the March 2026 Tashkent forum at all.

There is no separate water-use agreement between the Central Asian states and Afghanistan, and — with almost no government recognising the administration in Kabul — no realistic route to signing one. Bruce Pannier, The Times of Central Asia, reporting from Tashkent Water Week, 2026

03 / The landscape

What the canal is being cut through

The original Regional Landscape study assessed the canal's implications for Mazar-e-Sharif, Andkhoi and Sheberghan by examining how the region's ecosystems interact. That analysis holds, and it explains why the canal is not simply an irrigation question.

These cities sit on a broad deltaic plain at the confluence of the Band-e-Amir river system. The Balkh, Khulm, Sar-e-Pul and Shirin Tagab rivers create a natural boundary: fertile, rainfed or irrigated soils to the north, barren drylands to the south. The cold steppe climate — July averaging around 39°C, January around −2.1°C, with minimal rainfall — leaves the cities heavily dependent on both surface water and groundwater.

Two landscape types dominate. To the north, an extensive desert corridor of sand dunes runs parallel to the Amu Darya, contributing salt and dust that measurably affect public health. To the south, the Hindu Kush supplies the water and offers relief from summer heat. Significant cultural landmarks within a two-hour drive give the region genuine recreational value.

Landcover map of the Mazar region of Afghanistan showing the Qush Tepa Canal alignment — Landcover map of the Mazar region with the Qush Tepa Canal alignment. Source: FAO.

Four systems, and what the canal does to each

Principal landscape systems of the northern plain and their exposure to the canal
System	Current function	Exposure to the canal
The Amu (Amo) River	One of the largest relatively intact ecosystems on the northern border; supports diverse fish populations, recreation and local livelihoods	Direct abstraction at the Kaldar intake; downstream flow reduction in a river already at two-thirds of norm
The desert corridor	Arid and semi-arid belt from Jawzjan to Takhar; the principal source of seasonal sand and dust storms and of the region's strong winds	The channel bisects the desert, reducing contiguous habitat and cutting species movement corridors
Mountain-fed rivers	Khulm, Balkh, Sar-e-Pul and Shirin Tagab; delineate the boundary between cultivation and encroaching sand	Indirect relief — canal water is intended to reduce abstraction pressure, freeing year-round supply upstream
Wetlands	Scattered depressions and canal-margin accumulations acting as flood buffers, drought moderators and pollutant filters	Ambiguous: could be recharged by surplus canal water, or drained and converted to cropland where water is abundant

Classification adapted from the original Regional Landscape assessment (Afghan Architecture & Urbanism, 2025), with exposure ratings updated by Constructive Voices against 2025–26 construction and hydrological data.

The wetlands deserve particular attention, because they are the closest thing the region has to functioning Nature-Based Solutions already in place. They form in depressions where the water table is shallow, or where excess water accumulates along irrigation canals and runoff paths. They buffer floods, shorten droughts and filter pollutants — and they persist even in intensively cultivated zones, as part of a mosaic that includes irrigated fields, rainfed farms, shrubland and desert.

Map showing wetland locations across the northern region of Afghanistan — Wetland locations across the northern region — scattered through the agricultural mosaic rather than concentrated.

04 / The hazards

A hazard profile that got worse while the canal was being dug

Since 1980, natural hazards in Afghanistan have affected roughly 9 million people and killed more than 20,000. Flooding is the most frequent, with average annual damages around $54 million and major events exceeding $500 million. Earthquakes are rarer but deadlier — over 10,000 deaths since 1980 and around $80 million a year in damage. Droughts have affected 6.5 million people since 2000. Around 10,000 kilometres of road, some 15 per cent of the network including the Salang Pass, are exposed to avalanche risk.

Two events since the original report have sharpened that picture for the canal region specifically.

Hazard exposure of the three canal-region cities, with 2025–26 events
City	Drought	Earthquake	Flood
Mazar-e-Sharif (Balkh)	Moderate exposure; province prioritised in OCHA's 2026 drought framework	High — 12 deaths and 635 injuries in the November 2025 quake	Localised: southern high-slope and western marshy areas
Sheberghan (Jawzjan)	High; province prioritised in the 2026 framework	High	Lower
Andkhoi (Faryab)	Significant; province prioritised in the 2026 framework	Moderate	Lower

Baseline exposure ratings from the original Regional Landscape assessment. Casualty figures: OCHA and WHO situation reports, November 2025. Drought prioritisation: OCHA, Afghanistan Anticipatory Action Framework for Drought (2026), which names Balkh, Faryab, Jawzjan, Sar-e-Pul and Badghis.

On 3 November 2025 at 00:59 local time, a magnitude 6.2–6.3 earthquake struck near the Balkh–Samangan border, some 30 kilometres east-south-east of Mazar-e-Sharif, at a depth of 28 kilometres. Around 27 to 31 people were killed and more than 1,100 injured across six provinces. Several hundred homes were destroyed. The minaret of the Blue Mosque in Mazar-e-Sharif was badly damaged and cracks opened elsewhere in the shrine; the Bagh-e-Jahan Nama palace at Khulm was also hit. A mountain collapse in the Tangi Tashqurghan valley temporarily severed the Balkh–Kabul highway.

The second event is slower. OCHA's 2026 anticipatory action framework for drought prioritises five spring-rain and grain-dependent provinces — and three of them, Balkh, Faryab and Jawzjan, are the provinces the canal is being built to serve. The canal is, in that sense, being delivered into precisely the conditions that justify it and precisely the conditions that make it hardest to fill.

05 / The engineering

An unlined channel through sand

This is where the ecological and the technical arguments converge, and where the 2025 assessment's concerns have largely been borne out.

The canal bed is, over much of its length, unlined. Hydraulic engineers have warned since 2023 that without concrete or geotextile lining, a channel cut through sandy soil will lose a substantial share of its flow to seepage — the original design analysis put losses at up to 22 per cent over the first 40 kilometres alone. Sections have since been faced with concrete and stone, but the treatment is partial.

Seepage is not only a delivery problem. Water leaving the channel raises the local water table, and a rising water table in an arid basin with saline groundwater produces salinization. Northern Afghanistan is already among the country's high-salinity groundwater zones; close to a fifth of nationally monitored wells exceed WHO conductivity guidelines, with Mazar-e-Sharif specifically identified. The mechanism that turns new farmland into abandoned farmland is well documented in this basin — it is the story of the Karakum Canal, and ultimately of the Aral Sea.

The seepage lake

Satellite monitoring in late 2023 and 2024 identified a substantial body of water accumulating in the desert alongside the canal — reported at up to ten kilometres across at its furthest extent. Environmental monitors read it as a breach; Uzbekistan's space agency concluded the water had been deliberately diverted to manage seepage. Both readings point to the same underlying issue: the embankments are formed from excavated spoil heaped on either side, and in places that spoil is eroding in the wind.

The regional context is instructive rather than exculpatory. Roughly 65 per cent of irrigation canals across the basin lack concrete lining, and around 36 per cent of water is lost to the ground. Only about 30 per cent of Uzbekistan's canals have hard surfaces and around 20 per cent of Turkmenistan's. The Karakum Canal, operating since 1967, delivers only about half the water it withdraws — roughly 12 cubic kilometres a year, a fifth of the whole river, moved for a return of half. World Bank assessments cited in 2026 reporting find that only about half the water supplied in the lower Amu Darya reaches irrigated land at all.

Put plainly: the canal's projected losses are unexceptional by the standards of the basin it is joining. That is the problem, not the defence.

06 / The estimates

How much water — and why nobody agrees

Any reader following this story will have encountered wildly different figures for how much of the Amu Darya the canal will take. The spread is not carelessness; it reflects genuine uncertainty about seepage, about crop patterns, about whether the distribution network is ever finished, and about which flow year is used as the denominator.

Published estimates of Qush Tepa's abstraction from the Amu Darya
Source	Volume	Share of flow
Afghan officials (2026)	6–10 km³/yr planned	No more than 12% at full capacity
SIWI (2026)	—	8–20% of average annual flow
Crosslin, University of South Dakota (2025)	—	Up to 17%
Commonly cited regional estimate	—	~20%
Carnegie Politika (2025)	Up to 10 km³/yr	Around a third
Regional water-information consensus	Afghan basin use rising from 3–5 to 9–11 km³/yr	—

Compiled by Constructive Voices from published sources, 2024–2026. Percentages are not directly comparable: several are calculated against different baseline flow assumptions, and none of the higher estimates has been reconciled against the others.

The distinction that matters for downstream states is not the headline percentage but the timing. Peak abstraction coincides with peak irrigation demand, and in a drought year the same volume represents a far larger share. Analysts consistently note that a reduction of even 10 to 12 per cent would create serious difficulty for Uzbekistan's Khorezm region and Karakalpakstan, and for Turkmenistan's Dashoguz — the areas already closest to the edge.

Map showing the provinces covered by the Qush Tepa Canal — Provinces served by the Qush Tepa Canal: Balkh, Jawzjan and Faryab.

07 / For practice

Landscape-based measures that still make sense

As urban areas expand and the canal changes the region's water availability, natural resource management becomes the variable that determines whether the outcome is productive farmland or a second Karakum. Nature-Based Solutions offer a low-cost strategy that is already embedded in Afghan practice, addressing water security, pollution, hazard exposure, food insecurity and public health at once.

Two interventions are geographically specific and unusually high-value. Where the Balkh and flood rivers emerge from hillside canyons onto the Balkh plain — the points that feed both the cities and the agricultural land — there is scope for infiltration reservoirs that capture high-rainfall flows and bank them in aquifers for municipal use. And integrating treatment wetlands with existing wastewater systems would create biofilters north of the cities, in the saline desert, where the cleaned water could recharge shallow aquifers or irrigate a greenbelt against the dust.

Key recommendations

Interconnected agricultural networks. A continuous rather than fragmented agricultural landscape lowers local temperatures, improves resilience and gives wildlife uninterrupted habitat.
Sand dune barriers. Buffer zones between cultivated land and shifting dunes to slow desert encroachment.
Greenbelt development. Convert overgrazed rolling hills to greenbelts of local trees and shrubs — erosion control, air quality and recreation from the same investment.
Habitat connectivity. Wildlife corridors across the canal-divided desert, to preserve species movement and territory that the channel otherwise severs.
Water management infrastructure. Artificial dams and constructed wetlands in the south to manage upstream water, recharge aquifers and take pressure off drought years.
Wetland conservation and reforestation. Protect existing wetlands and link suitable areas into forested zones using salt- and drought-tolerant species, improving air, water and soil quality together.

For practitioners, the single most useful reframing is this: the canal is a conveyance project, and almost nothing about its current design addresses what happens to the water after it leaves the channel. Phase three — more than 500 kilometres of distribution canals — is the phase where salinization is either designed out or designed in, and it is the phase that has barely begun. Drainage, not delivery, is where the ecological outcome will be decided.

Map of Afghanistan's watershed zones — Afghanistan's watershed zones. The northern river basin, served by the canal, drains toward the Amu Darya.

08 / Conclusion

Two canals, not one

There is the canal Afghanistan is building and the canal the region is arguing about, and they are not quite the same object.

The canal Afghanistan is building is a domestic development project in a country where the FAO expects one of the world's largest humanitarian caseloads in 2026, where three of the five provinces flagged for drought anticipatory action are the provinces the canal serves, and where an entitlement to the river has been asserted, credibly, for decades without being exercised. On those terms it is difficult to argue that the project should not exist.

The canal the region is arguing about is a 20 per cent abstraction from a river running at two-thirds of normal, delivered through an unlined channel, into a basin with no mechanism that includes the abstracting state, at the end of a decade in which the flow has already fallen by 12 per cent.

Both descriptions are accurate. The gap between them is not a disagreement about values — Uzbekistan has conceded Afghanistan's right, and Kabul has asked for cooperation on riverbank stabilisation and water-saving technology. It is a gap in instrumentation and in governance: nobody is jointly measuring what the canal takes, and no forum exists in which the measurement would be binding if it were made.

That is what the built environment sector can actually influence. Not the diplomacy, and not the abstraction volume — but the drainage design, the lining specification, the wetland retention, the greenbelt, and the difference between 550,000 hectares of farmland and 550,000 hectares of salt.

Methodology and sourcing

Construction figures are as reported by Afghanistan's National Development Corporation and government spokespeople, cross-checked against independent satellite monitoring where available; they are not independently audited and should be read as claims rather than verified measurements. Hydrological figures are drawn from published regional reporting and academic sources listed below. Landscape and ecosystem analysis is retained from the original 2025 Regional Landscape assessment, with exposure ratings updated by Constructive Voices. Where this report classifies rather than cites — the landscape exposure table in section 03 — that is stated at the point of use.

Known data issues

The canal's dimensions are not consistently reported. Length appears as 285, 287 and 290 kilometres; width as 100 metres along the run and 125 metres at the intake; depth as 8 or 8.5 metres. These are all in circulation from official sources and the variance is likely definitional.
Abstraction estimates do not reconcile. See the table in section 06. The published range spans 6 to 10 cubic kilometres and 8 to over 30 per cent of flow. Anyone quoting a single figure should state the baseline it is calculated against.
Progress percentages are self-reported. Phase two moved from 81 per cent (December 2024) to 87 per cent (February 2025) to 90 per cent (May 2025) to 93 per cent (August 2025) to 95 per cent (October 2025) to 98 per cent (March 2026). The trajectory is credible; the precision is not.
Casualty figures for the November 2025 earthquake vary by source and reporting date, from 20 to 31 deaths and 320 to over 1,170 injuries as assessments progressed.

Primary sources

Asia-Plus / Azattyq Asia, Qosh-Tepa Canal: new water source or growing threat to Central Asia?, 10 August 2026
The Times of Central Asia, Afghanistan Advances Qosh Tepa Canal While Urging Regional Water Cooperation, 2026
The Times of Central Asia, The Amu Darya Stress Test, June 2026
Carnegie Endowment, Afghanistan's Qosh Tepa Canal Could Trigger a Central Asian Water Crisis
SIWI, Central Asia and Afghanistan: water cooperation at a critical juncture, 2026
OCHA, Afghanistan Anticipatory Action Framework for Drought (2026)
OCHA, Flash Update: Northern Region Earthquake, 3 November 2025
National Development Corporation, Phase Two of Qosh Tepa Canal Nears Completion, October 2025
E. L. Crosslin, Monitoring the Qosh Tepa Canal Project: A Geospatial Timeline, University of South Dakota, 2025
Faizee, Faqiryar, Mohamed & Sehring, The Qosh Tepa Canal: Implications for Amu Darya Hydrology and Regional Hydropolitics, 2026
Eurasianet, Unexplained spill fuels concern about Afghan canal project

The original Regional Landscape assessment on which this report builds — Transforming Landscapes: The Qush Tepa Canal and Northern Afghanistan's Ecosystems — was written by Abdullah Hakimi and first published by Afghan Architecture & Urbanism in March 2025. Maps reproduced with the original article; landcover map courtesy of FAO. Updated and restructured by Constructive Voices, August 2026.

Constructive Voices

Transforming Landscapes: The Qush Tepa Canal and Northern Afghanistan’s Ecosystems

Join Global Experts