As climate-driven glacier melt accelerates across the Himalayas, Nepal’s ambitious plan to lower water levels in four high-risk lakes after the deadly August floods faces sharp criticism for inadequate scale and incomplete risk reduction.

In response to the Nepal’s Bhotekoshi-Trishuli floods on 26 August, the Nepali government has accelerated a roughly $50 million project to lower the water levels of four potentially dangerous glacial lakes. Yet scientists who have studied these lakes for decades argue that the planned interventions, modelled on earlier limited efforts, will not make downstream settlements safe.

The catastrophic flash floods that tore through Nepal’s Bhotekoshi-Trishuli river corridor left more than 1,400 people dead and thousands still missing, exposing the lethal vulnerability of mountain communities to cascading cryosphere hazards. 

The project to drain the glacial lakes, formally known as Protecting Livelihoods and Assets at Risk from Climate Change-Induced Flooding in Glacial River Basins of Nepal and branded the GLOF Sanjeevani Project, targets Lumding Tsho and Hongu-2 in Solukhumbu, Lower Barun in Sankhuwasabha, and Thulagi (also called Dona) in Manang. These four lakes were flagged as high-risk in a 2020 assessment by the International Centre for Integrated Mountain Development (ICIMOD) and the United Nations Development Programme (UNDP). Funding comes primarily from a $36.16 million Green Climate Fund grant, with the government and partners contributing the balance. Officials say the initiative will benefit around 2.3 million people living in vulnerable mountain areas.

Limited Lowering Echoes Past Shortfalls

Nepal has already lowered two glacial lakes: Tsho Rolpa in Dolakha by about three metres in 2000, and Imja Tsho near Everest by roughly 3.4 metres in 2016. Both reductions were driven more by available budgets than by rigorous hydrological modelling. Scientists say the original plans called for far deeper drawdowns – up to 20 metres at Tsho Rolpa and at least 10 metres, preferably 20, at Imja – to meaningfully cut peak flood discharges.

“This idea of a three-metre reduction is based on Tsho Rolpa, and it is not sufficient,” said Alton C. Byers, a long-time specialist in Nepal’s mountain environments. “The original plan at Tsho Rolpa was to lower it by up to 20 metres, but for want of budget they stopped at three.” 

Jeffrey S. Kargel, who has studied Imja, Lower Barun and Thulagi, and Binod Parajuli of the Department of Hydrology and Meteorology (DHM) confirmed that budget constraints repeatedly overrode scientific recommendations. Even modest lowering reduces some risk, officials argue, but experts counter that spending tens of millions for only a few metres of drawdown achieves little downstream protection.

Project director Rajan Bhattarai has indicated that the four lakes will likely be lowered “somewhere around what was done at Tsho Rolpa and Imja.” Detailed design and environmental assessments are underway this fiscal year; actual construction of controlled drainage channels and gated structures will not begin immediately and could take up to six years. Workers face significant health risks from prolonged high-altitude labour as well as the constant threat of sudden flooding.

Beyond the Lake: Why Engineering Alone Falls Short

Lowering water levels is only one piece of the puzzle. Scientists emphasise that outlet regulation, stabilisation of moraine dams, comprehensive early-warning systems, and strict controls on downstream development are equally critical. Kargel noted that Imja now has a flow-regulation arrangement that helps manage both increased inflows and the natural downward migration of outlets as ice within the moraine melts. Similar measures proved useful at Tsho Rolpa.

Yet neither lake is fully secure. An ice avalanche or rockfall into a lake can still generate a wave that overtops the dam, regardless of modest water-level reductions. The 2017 Langmale event in the Barun valley illustrated the point: a relatively small lake received more than a million cubic metres of ice, rock and debris, producing a major flood. Size alone is a poor guide to danger.

Downstream exposure remains the larger gap. Hydropower plants, roads, bridges and settlements continue to be sited in high-risk corridors. “There is no sense in building expensive infrastructure, having a flood destroy it, building it again and watching it get destroyed again,” Kargel said. 

The August 2026 disaster itself was triggered not by a classic lake outburst but by a complex ice-rock avalanche that dammed and then released water – an event that closely resembled the 2021 Chamoli disaster in India’s Uttarakhand and earlier Nepali floods.

Thousands of Lakes, Sparse Knowledge

Nepal contains more than 2,000 glacial lakes. The 2020 ICIMOD-UNDP report identified 21 as potentially dangerous, eight of them newly recognised. Independent studies suggest the true number at elevated risk is considerably higher. Most lakes have never been surveyed in the field. Remote-sensing inventories provide area and expansion rates but rarely accurate depth or volume data needed for reliable flood modelling.

The new project includes a mandate to assess additional lakes and produce an updated risk report. Scientists insist this work must involve bathymetric surveys, field verification of moraine stability, and hydrodynamic modelling of low-, medium- and high-discharge scenarios. Only then can authorities decide whether to lower a lake substantially, reinforce its dam, install early-warning sensors, or simply restrict development in the flood path.

Rijan Bhakta Kayastha of Kathmandu University and postdoctoral researcher Nitesh Khadka both stress that exclusive focus on the largest lakes is a mistake. Small lakes can generate catastrophic floods when struck by mass movements, and glaciers themselves are becoming more unstable as temperatures rise and permafrost thaws.

A Regional Crisis Demanding Systemic Response

The Hindu Kush Himalaya is warming at roughly three times the global average. Glaciers are losing mass rapidly, lakes are expanding, and the frequency of complex cascading events is increasing. Nepal contributes a negligible share of global greenhouse-gas emissions yet bears a disproportionate burden of the impacts. International climate finance has been slow: the Green Climate Fund took years to approve the current project.

Similar pressures confront neighbouring countries. Bhutan, India and the Tibetan plateau all host growing inventories of potentially dangerous lakes. Cross-border river basins mean that an outburst or avalanche in one country can devastate communities downstream in another. Effective risk reduction therefore requires shared monitoring, data exchange and coordinated early-warning systems – areas where progress remains limited.

What True Safety Would Require

Experts outline a clearer path. First, conduct fresh, field-based assessments of all high-priority lakes rather than relying on outdated reports. Second, lower lakes substantially enough – often 10 metres or more – to reduce at least 70 per cent of potential downstream damage. Third, install robust, real-time monitoring and community-linked early-warning systems. Fourth, enforce land-use rules that keep critical infrastructure and settlements out of the most hazardous corridors. Fifth, integrate nature-based measures such as reforestation and bio-engineering with hard engineering.

Peru’s experience in the Cordillera Blanca, where dozens of dangerous lakes were lowered between the 1950s and 1970s, offers useful lessons in sustained, science-driven investment. Nepal’s previous projects at Tsho Rolpa and Imja demonstrated technical feasibility at extreme altitudes, but they also revealed the cost of under-ambition.