The wall of ice, rock and water that tore through parts of Nepal on the morning of August 26, killing several hundred people and leaving well over a thousand more missing across the Nepal-China border, has drawn attention back to a question India has been trying to answer for a decade: what equivalent risks lie on the Indian side of the Himalayas, and does the system built to track them capture the kind of event that unfolded?
India's surveillance of such risks has evolved since the 2023 Sikkim disaster and now covers hundreds of high-altitude water bodies across six states and Union territories, monitored by two separate agencies using two different methods. The event has now drawn attention to a distinct class of hazard does not involve a lake at all, but has to do with a mass of ice failing high on a mountain slope — one scientists had been flagging in the months before.
The lakes on India's list
Two national agencies are tracking Indian Himalayan glacial lakes in parallel. The National Disaster Management Authority (NDMA), which co-ordinates disaster preparedness at the central level and works through state disaster management authorities (SDMAs) for on-the-ground implementation, has identified 195 glacial lakes across Himachal Pradesh, Sikkim, Ladakh, Arunachal Pradesh, Jammu and Kashmir, and Uttarakhand that it considers vulnerable to a Glacial Lake Outburst Flood, or GLOF — the sudden release of water from a lake dammed by an unstable natural barrier of rock and debris left behind by a retreating glacier. Of these 195 lakes, 67 are classified Category A, or very high risk, and include Sikkim's South Lhonak Lake, Himachal Pradesh's Gepang Gath, and Bram Sar in Jammu and Kashmir; another 48 fall in Category B, or high risk, 74 in Category C, or moderate risk, and six have not yet been categorised.
{{/usCountry}}Two national agencies are tracking Indian Himalayan glacial lakes in parallel. The National Disaster Management Authority (NDMA), which co-ordinates disaster preparedness at the central level and works through state disaster management authorities (SDMAs) for on-the-ground implementation, has identified 195 glacial lakes across Himachal Pradesh, Sikkim, Ladakh, Arunachal Pradesh, Jammu and Kashmir, and Uttarakhand that it considers vulnerable to a Glacial Lake Outburst Flood, or GLOF — the sudden release of water from a lake dammed by an unstable natural barrier of rock and debris left behind by a retreating glacier. Of these 195 lakes, 67 are classified Category A, or very high risk, and include Sikkim's South Lhonak Lake, Himachal Pradesh's Gepang Gath, and Bram Sar in Jammu and Kashmir; another 48 fall in Category B, or high risk, 74 in Category C, or moderate risk, and six have not yet been categorised.
{{/usCountry}}Himachal Pradesh accounts for 48 of the vulnerable lakes, Sikkim 41, Ladakh 34, Arunachal Pradesh 32, Jammu and Kashmir 27, and Uttarakhand 13.
The Central Water Commission (CWC), under the Union Jal Shakti ministry, is responsible for the country's water-resources monitoring, and is doing a broader and mechanically different exercise. Its June 2026 monitoring report tracks 2,843 glacial lakes and other water bodies larger than 10 hectares across the Indian Himalayan river basins, of which 2,485 are glacial lakes; of the 681 glacial lakes that lie inside Indian territory, 147 recorded an increase in water spread during the month of June alone and were flagged for what the CWC called vigorous monitoring.
Together the two datasets give a broad picture of the high-altitude flooding risks.
Hazard on India's slopes
Alongside the lakes, a separate class of Himalayan cryospheric hazard is beginning to receive systematic attention, and it is one the Nepal event has brought sharply into public focus.
A study published in May 2026 in the journal npj Natural Hazards, led by researchers at the Indian Institute of Science, has produced the first basin-scale inventory of what glaciologists call hanging glaciers in India's Central Himalayas. A hanging glacier, as the name suggests, is an ice mass that sits high on the wall of a glacial valley — on steep slopes, often above icefalls or over cliffs — and is unlike a valley glacier that flows all the way down to a terminus, ends abruptly partway down the slope. These glaciers can remain in place on very steep terrain because parts of the ice are frozen to the underlying rock, an attachment that weakens with warming, meltwater intrusion, or the thinning that occurs as the ice loses mass. When that weakening progresses far enough, the ice can detach without warning, fall into the valley below, and pick up rock and sediment as it does so, forming the kind of ice-rock avalanche and downstream debris flow that struck Nepal.
The IISc study, focused on the Alaknanda basin in Uttarakhand's Garhwal Himalaya, mapped 219 such glaciers across an area of 71.7 square kilometres, containing an estimated 2.39 cubic kilometres of ice, of which 0.74 cubic kilometres is specifically the hanging mass — the portion whose failure would translate directly into a downstream hazard.
These glaciers sit on slopes with a mean surface angle of around 33.65 degrees, and some hanging masses are perched on inclines steeper than 50 degrees, at which point ice retention depends heavily on the ice-rock bond.
About 55% of the hanging glaciers the study mapped sit above larger trunk glaciers, meaning that in the event of a collapse the falling ice would land on more ice and its energy would be partly absorbed; the remaining 45%, however, overhang deglaciated or river valleys, where a collapse would fall directly onto exposed rock and river channels, with the risk to downstream infrastructure and communities depending on ice volume, velocity and how far the debris travels before stopping — what glaciologists call runout distance.
For 25 of these glaciers, the researchers ran collapse simulations, and the modelled results are what have given the study its policy weight. Simulated maximum flow heights reached 51 metres at Badrinath, 48 metres along the Badrinath-Mana road, and 40 metres at Hanuman Chatti — all of them sites on India's most heavily used pilgrimage routes.
The exposed population in the sample area was projected to rise from 380 people in 2000 to 8,540 by 2030. Anil Kulkarni, visiting scientist at the Divecha Centre for Climate Change at IISc and one of the study's authors, told HT the exercise was aimed at closing what he described as an under-studied gap in Indian glacier hazard assessment: “We realised hanging glaciers are not that well studied in the country and can pose an equally high risk. The assessment was done by us across Uttarakhand, to help agencies carry out mitigation measures.”
Also Read: Uttarakhand ramps up glacier monitoring after Nepal floods kill over 600
What India is doing about it
India's mitigation effort so far has been organised around lakes. Krishna Vatsa, member at the NDMA, told HT the core of the strategy is volume reduction — lowering the water level in a vulnerable lake so that the pressure on its moraine dam is reduced and, in the event of failure, the downstream impact is smaller. "We do this generally by gradually pumping out the water, so that the pressure is reduced, but even that has its own challenges because of the height and terrain," he said.
"Solar-based pumps are used, as there is no electrical connection, but even then, solar energy is not always available." Most of the lakes in question sit above 4,000 metres, in terrain without roads or grid power, with narrow weather windows for any physical intervention. But the technique has been used at scale before, most notably at South Lhonak, where a siphoning effort using high-density polyethylene pipes reduced the lake's water volume by about half in the years before 2023. That lake was the site of a GLOF in October 2023 that killed dozens downstream in Sikkim and destroyed the 1,200MW Teesta III hydropower project.
Alongside volume-reduction work, NDMA is testing early-warning systems, or EWS, designed to give downstream communities time to move to safety in the event of a lake failure — typically a combination of water-level sensors upstream, satellite-linked telemetry, and downstream alert infrastructure. "We recently did some work in Himachal Pradesh, where EWS was tested at a lake, but at a lower height," Vatsa said. “We will now test this at lakes at much higher heights, particularly in Sikkim. Once tested, it can be replicated across all dangerous lakes.”
He added that on-the-ground reconnaissance had been substantial, with at least 30 physical expeditions carried out over the past several months to assess terrain, moraine stability and outlet conditions at high-priority lake sites.
The institutional architecture involves a partnership between NDMA and the state disaster management authorities, with technical institutions in a supporting role. VK Gahalaut, director at the Wadia Institute of Himalayan Geology, told HT the identification exercise had accelerated after the October 2023 South Lhonak disaster, and that his institute's role was to ensure coordination between agencies and to provide scientific inputs. "The process to identify such lakes was carried out by NDMA, which entrusted this job to the respective SDMAs — to identify lakes where GLOF is possible," he said. "This picked up particularly after the 2023 Sikkim floods." He added that on-ground work at identified lakes had already begun.
Why the challenge is hard
The measures described above address the lake population reasonably well, and state capacity to identify vulnerable lakes and act on them has grown meaningfully in the years since the 2023 South Lhonak Lake outburst flood. But the lake at South Lhonak had been on India's risk radar for a decade before it failed — an article by scientists from the National Remote Sensing Centre in the journal Current Science had flagged its very high risk of a GLOF as far back as 2013 — and had, by 2023, been the target of active mitigation, as Vatsa explained.
The October 2023 failure still occurred, because the trigger was not a slow overflow of the moraine dam but a sudden collapse of the lake's northern lateral moraine, which sent an impulse wave over the barrier and initiated the outburst. Volume reduction, in other words, addresses one class of GLOF — the slow-onset kind driven by dam weakening or overflow — but not the kind triggered by something falling into the lake from above.
The Nepal flash flood widens that limit further, because it originated in a class of event the lake list does not directly capture at all. The initial failure was not of a lake but of a mass of high-elevation ice on a slope; the lake in that story came afterwards, downstream, formed when the debris flow blocked the Lhende Khola and pooled water behind the blockage. This is what glaciologists mean when they describe such events as cascading hazards, in which one failure triggers the next in a sequence that can play out within hours.
The IISc study's mapping of 219 hanging glaciers in the Alaknanda basin alone, and its identification of high-slope, ice-cliff configurations in areas above densely-visited pilgrimage sites, points to a hazard India has only begun to inventory, and one for which the mitigation methods used on lakes offer no direct equivalent.
Reacting to the Nepal event on Sunday, Simon Stiell, executive secretary of the UN Framework Convention on Climate Change, said the sudden glacial collapse and flash flood was "again a devastating reminder of how fragile these mountain environments can be, particularly in the Hindu Kush Himalayas," and pointed to the wider forcing at work. "We also know a warming climate, driven by humanity burning colossal amounts of coal, oil and gas, is making tragedies like this one, and so many others around the world, much more likely, frequent and severe, with vulnerable communities often hardest hit," he said. "But the solutions to global heating are equally clear — especially a faster global shift from fossil fuels to renewables and more investment in climate resilience."