Skeleton Lake: How DNA Cracked Roopkund's Himalayan Mystery

High in the Garhwal Himalaya, above the tree line and well inside the kind of cold that cracks lips and numbs fingers, sits a small glacial pool that locals and trekkers know by an unsettling nickname: Skeleton Lake. Its real name is Roopkund, and when its thin sheet of ice melts each year, the shallow water and the rubble around its edge reveal hundreds of human bones—skulls, femurs, ribs, even hair and fragments of leather still preserved by the freezing air. For decades, the obvious question hung over the site like the surrounding peaks: who were these people, and what killed them? In 2019, an international team of scientists finally pried open the answer using ancient DNA—and what they found was stranger than any of the legends.

A lake of bones at 5,020 metres

Roopkund lies in Chamoli district, Uttarakhand, perched at roughly 5,020 metres (about 16,470 feet) in the lap of two towering summits, Trishul and Nanda Ghunti. It is not large—you could walk its circumference in a few minutes—and for most of the year it is frozen over. The skeletal remains scattered in and around it are thought to represent more than 300 individuals, a staggering concentration of human dead for such a remote, hostile spot with no permanent settlement anywhere nearby.

The remains were brought to wider attention in 1942, when a forest ranger named Hari Kishan Madhwal, patrolling the Nanda Devi sanctuary, came upon the grim scatter of bones. Wartime India briefly entertained a darker fear—that the dead might be soldiers, perhaps a Japanese force trying to slip through the mountains—before that idea was dismissed. The bones were plainly old. But how old, and whose, would take another lifetime to establish.

The legends people reached for first

Nature abhors a vacuum, and so does the human imagination. Before science could speak, folklore filled the silence. The most enduring local story tells of Raja Jasdhaval, a king of Kanauj, who set out on a pilgrimage toward the shrine of the goddess Nanda Devi accompanied by his pregnant wife, servants, dancers and musicians. The procession was lavish and, the legend insists, improperly so for such sacred ground. Enraged by the noise and the defilement of her mountain, the goddess is said to have unleashed a storm of hailstones "hard as iron," annihilating the entire party near the lake.

That tale is woven into a living tradition. Every twelve years, devotees undertake the Nanda Devi Raj Jat, an arduous pilgrimage through this high country to honour the goddess, a route that passes close to Roopkund itself. The story of a doomed royal procession, in other words, is not a museum piece; it is part of how people in the region still understand their landscape.

A surprisingly modern clue: death from above

The folklore turned out to contain a kernel that science could test. A forensic examination led by anthropologist Subhash Walimbe, reported in the mid-2000s, noted a striking pattern in the injuries. Many of the dead had suffered short, deep cracks to the top of the skull and to the shoulders—wounds consistent with heavy, rounded objects striking from directly overhead, rather than weapons, falls, avalanche or landslide. Crucially, the rest of the skeletons were largely unmarked. There were no defensive wounds, no signs of a fight.

The interpretation that emerged was almost cinematic: a sudden, violent hailstorm pelting an exposed group with ice balls the size of cricket balls, with nowhere to shelter on the open mountainside. It was a tidy fit for the legend of hailstones flung by an angry goddess. For a while, it looked as if the mystery of Roopkund was essentially solved.

DNA blows the case wide open

Then came genetics, and the neat picture shattered. In a study published in Nature Communications in 2019, a team of 28 researchers from institutions across India, the United States and Europe recovered genome-wide ancient DNA from 38 of the Roopkund skeletons, combining it with radiocarbon dating, stable-isotope analysis of ancient diets, and physical examination.

The bones did not belong to one tragedy. They sorted into three genetically distinct groups. The largest—23 individuals—carried ancestry that sits comfortably within the range of present-day South Asians, though their genetics suggested they came from varied backgrounds rather than a single community. A second group of 14 people had ancestry typical not of India at all, but of the eastern Mediterranean, the region of present-day Greece and Crete. And a single individual carried ancestry linked to Southeast Asia.

The dating was the real shock. The South Asian group died around the 7th to 10th centuries CE, apparently across more than one event rather than in a single instant. But the Mediterranean and Southeast Asian individuals died roughly a thousand years later, around the 17th to 20th centuries. Two clusters of dead, separated by a full millennium, had come to rest in the same tiny pool.

What we now know—and what we don't

This finding quietly demolished the idea of a single catastrophic afternoon. A hailstorm might plausibly explain a group of pilgrims caught in the open around 800 CE—the injury evidence and the early date can sit together. But it cannot explain a second, genetically unrelated population arriving and dying near the lake some thousand years afterward. Whatever happened at Roopkund happened more than once, to very different people.

The study also closed off some tempting explanations. Stable-isotope dietary signatures differed between the groups, consistent with their distinct origins. The researchers found no trace of bacterial pathogens that would point to an epidemic, and the absence of trade goods or weapons argues against the dead being a battle party or a merchant caravan wiped out together. As for the eastern Mediterranean group, the most arresting question of all remains open: what were more than a dozen people of Greek-island ancestry doing, and dying, on a Himalayan slope two hundred years or so ago? No documentary record has been found to explain their presence. The genetics are certain; the history behind them is a blank.

Why a frozen pool matters

Roopkund is a vivid lesson in how science reshapes a story we thought we understood. For generations the lake was a single ghost story—a king, a goddess, a storm. Careful forensics seemed to confirm it. Then a more powerful tool revealed that the truth was layered, literally, with bodies from epochs that never met. The romance of one tragic legend gave way to something harder and more interesting: evidence of repeated, still-unexplained human journeys to one of the least hospitable places imaginable.

It is also a reminder of how much information survives in old bones, and how much can be lost. The remains at Roopkund have been disturbed for decades by trekkers who have moved, stacked and pocketed them, scrambling the very context archaeologists need. The site sits within a fragile high-altitude ecosystem near Nanda Devi, and the bones are an irreplaceable archive. Future analysis—of ancient proteins, of more individuals, of the leather and wood preserved in the ice—may yet name the routes these travellers followed and the reasons they came. For now, Skeleton Lake keeps most of its secrets, having surrendered just enough to prove that the real answer was far stranger than the legend it replaced.

Source: nature.com