The wall of water gave them almost no time.
It tore down from the high Himalayas carrying ice, rock, mud and uprooted trees, entered the Bhote Koshi and Trishuli river system, and swallowed villages, roads, bridges and hydropower installations along the Nepal–Tibet frontier. Water levels reportedly rose by as much as nine metres downstream within half an hour.
By early Thursday, police had recovered at least 157 bodies, according to The Guardian’s live reporting. Hundreds remained missing, including pilgrims, tourists, hydropower workers and residents of settlements that had stood in the flood’s path.
The figure is already appalling. But in the history of sudden floods—events in which lakes rupture, dams fail or mountains abruptly send water and debris down narrow valleys—it is not yet among the very worst.
Floods are often discussed as though they belong to one category, but a river rising over several days is fundamentally different from a wall of water arriving in minutes. The World Meteorological Organization defines a flash flood as a sudden event in which fewer than six hours separate the observable cause from the flooding.
Such floods cause more than 5,000 deaths globally each year and have the highest mortality rate among flood types, according to the WMO.
The Nepal–Tibet disaster appears more violent still. Satellite analysis cited by Reuters indicated that part of a glacier at about 5,200 metres may have collapsed into a valley more than 1,200 metres below. The impact released a torrent of ice, sediment and boulders.
Scientists have described it as an ice-rock avalanche rather than a conventional rain-induced flood, although the precise sequence remains under investigation. Reuters reported that Nepal has lost nearly a third of its ice in three decades, with melting accelerating sharply over the past 10 years.
History’s deadliest comparable disasters show what happens when enormous volumes of water meet populated valleys and little warning.
In December 1999, weeks of rain saturated the mountains above Venezuela’s Caribbean coast. The slopes then disintegrated. Thousands of landslides converged with flash floods and debris flows, burying communities in Vargas state. The death toll has never been established with certainty.
The US Geological Survey places it at roughly 19,000, while broader estimates range between 15,000 and 30,000. Entire neighbourhoods disappeared beneath sediment deep enough to redraw the coastline.
Two decades earlier, in August 1979, the Machhu-II dam failed above Morbi in the Indian state of Gujarat. Water reportedly rose by as much as nine metres in parts of the city within minutes. Nobody knows exactly how many people died.
Estimates range from 1,800 to 25,000—a discrepancy that illustrates another feature of sudden disasters: they can destroy not only people and homes but the records needed to count them. The Association of State Dam Safety Officials say the true toll remains unknown.
The most lethal dam-collapse disaster came in China in 1975. Typhoon Nina brought extraordinary rainfall to Henan province, overwhelming Banqiao and dozens of surrounding dams.
A massive flood wave raced across the plains. About 26,000 people were reported to have died directly in the flooding; estimates reaching 240,000 include those who subsequently died from famine and disease.
Banqiao does not fit neatly beside a glacial collapse in the Himalayas. It was a cascade of engineering, forecasting and administrative failures on a vastly greater scale. But it demonstrates how natural extremes become human catastrophes when infrastructure is poorly designed, warnings fail and people are settled in the path of water that has nowhere else to go.
The closer historical parallel to Nepal may lie in Peru.
In 1941, Lake Palcacocha, held behind unstable glacial debris high above Huaraz, suddenly released millions of cubic metres of water. The flood descended roughly 1,500 metres towards the city, destroying about a third of it. Estimates of the dead range from at least 1,800 to 5,000; NASA uses a figure of about 4,000.
The danger has not vanished. The lake is now far larger because the glacier feeding it has retreated. Huaraz, meanwhile, has grown from a small provincial city into an urban area of more than 100,000 people. The hazard above the city and the number of people below it have expanded together.
That is also the Himalayan problem.
The mountains are warming, glaciers are retreating and unstable lakes are forming or growing. At the same time, roads, dams, hotels, border facilities and hydropower projects are advancing farther into steep valleys. Climate change is altering the source of the danger; construction is increasing the number of people and assets exposed to it.
The 2013 Uttarakhand disaster demonstrated the consequences. Extreme rain, flash flooding and landslides devastated the Kedarnath region of northern India during the pilgrimage season.
More than 5,700 people who remained missing were eventually presumed dead, while later assessments placed the overall toll at about 6,000. Roads and bridges collapsed, leaving hundreds of thousands stranded. The Asian Disaster Reduction Center records an estimated 5,748 deaths.
Ten years later, a glacial-lake outburst in Sikkim sent a surge down the Teesta River, killing 179 people and destroying bridges, settlements and part of a hydropower dam. In Nepal itself, repeated floods have damaged communities and cross-border infrastructure with unsettling frequency.
—