Twenty-One Glaciers: The Number Nobody Reported

When the abridged version of A Resilient Himalaya was released on 9 September, the coverage converged within about forty-eight hours on a single line: Himalayan glaciers are losing mass 65 per cent faster than they were a decade ago. Reuters led on it. Down To Earth led on it. So did most of the Indian business press, and most of the international wires that picked the story up.

It is a good number and it deserved the attention. But there is another figure in the same report, and as far as we can tell it went almost entirely unreported.

There are roughly 40,000 glaciers in the Hindu Kush-Himalaya. About 21 of them are regularly monitored.

That is one-tenth of one per cent, and it is the finding that makes all the others difficult to act on. The report, produced by Systemiq with the Integrated Mountain Initiative, supported by The Rockefeller Foundation, with technical input from ICIMOD and the G.B. Pant National Institute of Himalayan Environment, puts the full version out in October. We would like the monitoring gap to be in the headlines that time.

The size of the thing being measured

The same report states, with more institutional weight behind it than the claim has previously carried, that the Himalayan system underpins more than 20 per cent of India's GDP. Irrigation water, hydropower, industrial cooling, municipal supply, and a tourism economy that recorded over 400 million visits in 2025 all depend on it.

The region covers about 18 per cent of India's land area and absorbs roughly 35 per cent of the country's natural disasters.

The states that host it capture something like 5 per cent of the economic value it generates.

Put those figures beside each other and the Himalaya stops reading as an environment and starts reading as infrastructure: a system carrying a fifth of national output, maintained by whichever hill state happens to contain the broken section, instrumented at a rate that no other asset class in India would accept. A port authority knows its berth depths. A distribution company knows its transformer loading. A fifth of GDP rests on 40,000 ice bodies of which we watch 21.

What the gap costs in practice

There is a little over 40 GW of hydropower operating across the Himalayan arc, and something in the region of 200 GW planned. Work on the Third Pole identifies 105 operational or planned projects exposed to glacial lake outburst floods. Each of those was financed against a hydrological record that assumed a stationary climate, which is to say that lenders and insurers are underwriting twenty-five-year concessions on twentieth-century inflow curves.

India has assessed 189 high-risk glacial lakes and classified 56 of them as very high risk. That assessment is a screening exercise rather than a monitoring system, and the distinction is not pedantic. Screening tells you where to look. Monitoring tells you when to move. Without the second, mitigation money follows the politics rather than the hazard, which is more or less what has happened.

There is also a quieter cost. Mountain states arguing for a larger share of central transfers have no instrumented basis on which to quantify either the service they provide or the risk they carry. The 5 per cent value-capture figure is a debating point. It is not yet an invoice, and it cannot become one without measurement.

August supplied the evidence

At around 8.40 in the morning on 27 August, a glacier failed at roughly 5,200 metres on the Nepal-Tibet border. Ice and rock dropped some 1,200 metres into the Lhende, dammed it, and the dam gave way.

The toll has climbed through September. Reporting on the day put it near 390 dead with more than 1,500 missing across Rasuwa, Nuwakot and Dhading in Nepal and Gyirong county in Tibet; by the middle of the month the figures being carried were over 1,300 confirmed dead and more than 5,300 missing. Nineteen bridges went. Around 40 kilometres of road. Twelve hydropower plants were damaged, taking roughly 400 MW out of service, the Upper Trishuli project among them, and the assessed damage came to about $850 million, or 130 billion Nepali rupees. Analysts put the delay to Nepal's hydropower build-out at five to six years against a target of nearly 30 GW by 2035.

One glacier, unmonitored, removed half a decade from the delivery schedule of a country's principal export industry.

Whether this can happen on the Indian side is not really in question; Uttarakhand has answered it more than once. The question worth asking is whether any of the 105 exposed projects would have received more warning than Upper Trishuli did. On current instrumentation, they would not.

The one lever that is genuinely local

Most Himalayan climate argument ends up in the same cul-de-sac. Emissions are global, India's historical contribution is small, the mountain must adapt and that is that. The report's most useful contribution is identifying where that logic does not hold.

Roughly a third of Himalayan glacier melt is driven by black carbon, the soot from incomplete combustion that settles on ice, darkens it and accelerates absorption. Industrial sources account for somewhere between 32 and 42 per cent of human-caused black carbon emissions in the region.

Black carbon stays in the atmosphere for days or weeks rather than centuries. It is emitted regionally and it deposits regionally. Reducing it therefore produces a measurable local benefit on a timescale a state government can actually be held to, which is not true of carbon dioxide and is the reason brick kilns, diesel generation, open burning and domestic solid fuel belong in a Himalayan economic strategy rather than only in an environmental one. Across the ten transitions it sets out, the report puts the adjusted return at around eight rupees of benefit for every rupee spent.

Before October

Moving from 21 monitored glaciers to a few hundred, with automated weather stations, lake-level telemetry and satellite-fed early warning on the 56 very-high-risk lakes, is capital expenditure in the low hundreds of crores. Set against a fifth of GDP, and against an $850 million single-event loss immediately next door, that does not need a cost-benefit case so much as a decision and a line item.

It also needs to be public and machine-readable, which is the part most likely to be skipped. Monitoring data that sits inside a ministry cannot be used by an insurer to write a parametric product, by a lender to size a debt service reserve, or by a state to make its case to a Finance Commission. Open cryospheric and hydrological data is the precondition for private capital taking Himalayan risk at all, and at present there is almost nothing for it to take a position on.

Up to 80 per cent of the present volume of Himalayan glaciers may be at risk by the end of the century. That is a horizon long enough to invite postponement. But the decisions being taken now, on 200 GW of planned hydropower and on twenty-five-year concessions, are being taken against a river system almost nobody is measuring, and those decisions will not wait for October either.