Five of Kashmir's Glacial Lakes Are at ‘Very High Risk’ of Bursting

SRINAGAR: Months before a Himalayan glacier collapsed killed more than 1,300 people in Nepal, roughly 1,200 kilometres away, a peer-reviewed study had already mapped and ranked all 155 glacial lakes in the Jammu and Kashmir Himalayas, and found five of them at “very high” risk of a glacial lake outburst flood, or GLOF.

Published in January in the Journal of Glaciology, the study went largely unnoticed outside scientific circles, barring some brief news reports.

What's at Stake

Together, the five highest-risk lakes threaten 2,704 buildings and 15 major bridges within their mapped flood paths. This figure has already been cited by the J&K government in its own state assembly. That count doesn't extend to population. A single structure in these high-altitude valleys can house an extended family, besides the season's force of migrant labourers, so the human toll of any actual outburst would likely run higher.

The single largest exposure comes from Nundkol and Gangabal, a connected pair of lakes whose combined outburst would threaten 1,184 buildings, four bridges and a hydropower plant. This is more than either of the other two clusters. Bhagsar follows closely, endangering 1,114 buildings, six bridges and the Mughal Road, a key alternate route across the Pir Panjal range.Bramsar, which the study ranks as the single most likely of the five lakes to actually fail, sits at the other end of the exposure scale. Paired with Chirsar, it threatens the smallest area of the five - 406 buildings and five bridges.

How the Risk Was Mapped

To arrive at this ranking, researchers at the University of Kashmir, with colleagues at IISER Pune and NIT Srinagar, combined 32 years of satellite imagery, from 1992 to 2024, with a structured scoring method that weighs ten factors known to precede real GLOFs elsewhere in the Himalaya. These factors include a lake's size and how fast it is growing, what its natural dam is made of, the steepness of the surrounding slopes, whether other lakes drain into it in a chain, and whether it sits near active faults or thawing permafrost.

To check that the method actually works, the researchers tested it against eight lakes elsewhere in the Himalaya that have genuinely burst since 2010, including South Lhonak in Sikkim, whose 2023 outburst killed 178 people and destroyed three hydropower plants. Every one of them was correctly flagged as very high risk.

Five Lakes, Five Different Reasons

Ranked by susceptibility, the five lakes are Bramsar, Nundkol, Bhagsar, Chirsar and Gangabal and the factors for risk were distinctive for all.

Bramsar and Chirsar sit in direct contact with their parent glaciers and are growing fast. Bramsar's surface area has expanded 43% since 1992, Chirsar's by 65%. Both are held back by loose debris dams below slopes as steep as 62 degrees. The study flagged this combination as especially prone to failure by landslide or ice avalanche, the same mechanism suspected in several of the Himalaya's worst historical GLOFs.

Nundkol and Gangabal sit further downslope and are unusually large and elongated, and each is fed by smaller lakes upstream, forming a “cascade” structure meaning a smaller outburst higher up could overtop and trigger a larger one below. There is one mitigating factor.

Gangabal's dam is mostly solid bedrock rather than loose debris, and a J&K government field survey published this year independently found it “currently stable”, which the study's authors say, makes a full dam collapse there less likely than the worst-case exposure numbers imply, even as smaller, cascade-triggered overflows remain possible.

Bhagsar, the fifth, doesn't touch a glacier at all. But it is large spanning 76 hectares of land and sits beside degrading permafrost, and lies in one of the most seismically active stretches of the Himalaya.

Ice-Contact Lakes Are the Ones to Watch

The underlying driver is familiar. Kashmir's glaciers are thinning faster than most of the rest of the Himalayan arc, and the lakes that form directly against melting ice or ‘ice-contact lakes’like Bramsar and Chirsar are growing to match. Over the 32 years the study covers, ice-contact lakes expanded by 26% in area; lakes sitting further from the ice barely moved at all.

No GLOF has yet struck the Kashmir valley itself. That absence of a precedent is precisely why the researchers frame their findings as an early warning rather than a postscript.

Efforts on Ground

J&K's Focused GLOF Monitoring Committee, chaired by the Principal Secretary of the Home department, has run field expeditions to high-risk lakes including Gangabal, and to three lakes in Kishtwar district's upper Chenab catchment - Mundiksar, Hangu, and a third still unnamed. Last year's state budget set aside Rs 15 crore for GLOF monitoring and early-warning systems, and the government has separately ordered a high-precision robotic echo-sounding boat to carry out bathymetric surveys for measuring the actual depth and volume of the riskiest lakes for the first time, rather than relying on surface area alone, as the current study necessarily does.

Authorities say at least 14 high-risk glacial lakes across the wider Union Territory are now under watch, a list that extends beyond the Kashmir valley into the Chenab valley, which several glaciologists consider even more exposed than Kashmir itself.

Days after the Nepal disaster, the Union Home Secretary summoned officials from J&K and four other Himalayan states to New Delhi to review this preparedness. It is a reminder, more than a trigger, of work already under way.

What the Researchers Recommend

The study's calls for continuous satellite and field monitoring of the five highest-risk lakes, engineering work to stabilise dam material where feasible, community-level early warning systems that can alert downstream villages within minutes rather than hours, and keeping new construction out of the mapped flood-impact zones altogether.

A prototype early-warning system built by India's C-DAC and already tested in Himachal Pradesh's Lahaul-Spiti shows the underlying technology exists. The harder task is making sure a warning reaches people with enough time to matter, in terrain where evacuation routes are few and warning windows are short.

The researchers are candid about the limits of their own method. It rests on expert judgement and satellite imagery rather than physical surveys of every dam, and its exposure figures should be read as indicative rather than exact.

But as a first map of where the danger is concentrated, in a region that has never recorded a glacial lake outburst and cannot afford to wait for its first one to arrive, it is significant and merits full attention.