A devastating flood last week on the upper reaches of the Rishi Ganga River, in India’s northern state of Uttarakhand, is a stark warning of the hazards of hydropower development under the dual threats of climate change and young, unstable mountains, scientists said last Friday in a webinar hosted by the The Earth Institute at Columbia University, New York, US.
The flood came thundering down the Alaknanda River, a Ganges River headstream, and hit two dams (one still under construction), smashing both and killing dozens of people, including construction workers caught unawares at the second dam site.
“Somewhere around 200 people are still missing,” said Joydeep Gupta, South Asian director of The Third Pole, an organisation dedicated to compiling science news about the region’s glaciers.
Initially, scientists assumed the cause was a glacial lake outburst, something that happens when rapidly melting glaciers produce short-lived lakes.
“Typically, these are deep and sediment loaded, in very, very steep topography,” says Joerg Schaefer, a climate scientist at Columbia’s Lamont-Doherty Earth Observatory, who was not part of the webcast. “If you have an event that destabilises a lake, then this sediment and water funnels down and gets accelerated by gravity.”
The result: a dam-busting flash flood.
But when scientists looked at satellite images from before the flood, Schaefer says, they found no such lake. “If there was no lake, you cannot have had a glacial-lake outburst flood,” he says.
Instead, says Earth Institute climate researcher Benjamin Orlove, it appears to have been caused by a giant landslide that started high up on a mountainside, then plummeted nearly 2,000 metres downslope in a mix of rock, ice, and water. “Just imagine dropping a chunk of a mountain two kilometres,” Orlove says.
What happened then is open to debate. One possibility is that avalanche debris contained enough water that it continued to flow once it hit the valley.
But, Schaefer says, it’s also possible it temporarily blocked a river. “The water level rises, the pressure on the landslide gets higher and higher, then the landslide gives way and you have this wave of water,” he says.
Scientists have for years warned about the risks of glacial-lake outburst floods and the hazards of building dams downstream of retreating glaciers. Which makes it ironic that the devastation, when it hit, didn’t come from a lake.
But, Schaefer says, that doesn’t mean there’s no threat from glacial lake outbursts, because climate change is rapidly multiplying the number of these lakes and their concomitant risks. “If we had tens of them ten years, ago, now we have hundreds,” he says.
Furthermore, he says, even this flood is almost certainly related, at least in part, to climate change and global warming.
That’s because climate change can melt permafrost—underground ice that can hold steep slopes of jumbled rocks together. “If you start making it warmer, part of that ice melts and you lose that glue,” he says.
An obvious solution is to heed scientists’ warnings and quit building dams in regions where unstable mountains and climate change can destroy them, whether by glacial lake outbursts, landslides, or some other mechanism (such as an earthquake).
But that’s not likely to happen, says Ramesh Bhushal, South Asia coordinator of the Earth Journalism Network. In part, he says, it’s an issue of “gigantism,” in which each developing country wants something enormous to show off.
Another solution, Schaefer says, is to develop warning systems that can give advance notice of landslides and glacial-lake outburst floods. The rumbles from such events are so intense, he says, that it’s possible to trace them with seismometers as much as 100km away.
In a paper in the 16 September 2020 issue of Science Advances, for example, Schaefer and graduate student Josh Maurer (as well as others) were able to use old seismology records to follow the progress of a 1994 glacial-lake outburst flood that killed 21 people in Bhutan.
But modern technology is developing an even simpler form of warning, one that probably saved lives in India. People upstream saw what was happening and went live on Facebook, Twitter, and WhatsApp, Gupta says, posting messages that were read by people downstream.
Not that everyone was spared, but many probably were. “Early warning people to people did work,” he says.