'The lakes up there were already at the brink due to accumulated rainwater, and then snowmelt occurred.'
'When one lake breached, it triggered a cascading effect -- other lakes breached subsequently.'
In this first part of a two-part interview with Prasanna D Zore/Rediff, Dr Pradeep Srivastava, Associate Professor at the Department of Earth Sciences, IIT Roorkee, explains why Tuesday's flash floods near Kheer Ganga in Uttarakhand's Dharali village in Uttarkashi were not caused by a cloudburst, but by a series of glacial lake breaches triggered by continuous rain and snowmelt.
Dr Srivastava, who till 2021 worked as a scientist at the Wadia Institute of Himalayan Geology based in Dehradun, Uttarakhand, breaks down the mechanics of the destructive debris flow that followed, the role of unstable terrain above 2,200 metres, and how poor land use and construction in older riverbeds made the disaster worse.
Can you explain what happened at Kheer Ganga and whether this qualifies as a cloudburst according to meteorological definitions?
According to the available news and data from IMD (Indian Meteorological Department), the definition of a cloudburst is 100 millimetre of rainfall in a very small time -- for example, in one hour or less.
According to that standard, the rainfall there was around 40mm, and that too spread over 24 hours. So this rainfall doesn't qualify as a cloudburst by meteorological definition.
However, there are 23 other parameters that need to be considered. The Kheer Ganga catchment extends to higher elevations where there are snow caps and small glaciers. There are also lakes there.
The Wadia Institute of Himalayan Geology has marked several lakes in the Ganga catchment -- around a thousand plus lakes. So there are lakes in the Kheer Ganga catchment as well.
What role did these glacial lakes play in the Dharali flash floods?
We were reviewing a documentary by Professor Chunilal from Srinagar, where he pointed out specific lakes in the upper catchment of Kheer Ganga. Now, two critical factors came into play simultaneously: Continuous rainfall was occurring, and there's snow on steep slopes, plus there are multiple lakes at higher elevations.
This continuous rainfall slowly saturates the soil with water. The lakes up there were already at the brink due to accumulated rainwater, and then snowmelt occurred.
When one lake breached, it triggered a cascading effect -- other lakes breached subsequently. The soil cover on the slopes below was also saturated at this time that all of that material -- water, soil, rocks, boulders -- came down through a very steep gradient system in what we call debris flow.
Can you explain the mechanics of debris flow and why it's so destructive?
Debris flow is a phenomenon where while water has a density of one gram per cubic centimetre, debris flow becomes almost double that density. So its thrust increases dramatically, and due to this thrust and increased velocity, it can transport large boulders downstream.
The thrust of such flows is enormous. Now, the old course of Kheer Ganga had several buildings and colonies developed on it because that course had become inactive and was slightly elevated. People and planners thought it was now a safe place (to build houses and undertake construction activities).
But it's difficult to determine exactly when this old course changed. We can see evidence that both old and present courses have been active historically. In the 1700s, there were 23 major events in this region. At that time, Kheer Ganga had adopted its present course, but yesterday's event followed the older, original course.
Now, no matter how strong a structure you build in these old courses, it cannot withstand the thrust of such flows. It simply cannot. This is the nature of such events -- they're natural disasters.
The key was understanding the landscape properly. The older villages -- their locations are more what we call 'informed.' There was no damage to those traditional settlements. But these newer settlements that developed in the old river course got damaged completely.
How many such vulnerable spots exist across Uttarakhand where old river courses have changed?
In Uttarakhand, particularly in areas like Chamoli and at such elevations, there's what we call the Main Central Thrust.
In these areas, hill slopes rise very dramatically -- the gradient becomes extremely steep around 2,500 metre elevation. Above 2,000 metre elevation, the gradient of hill slopes is consistently very steep.
On top of those peaks, there's also permafrost, ice caps, and small glaciers. A similar event happened in 2013 at Kedarnath, but at a larger magnitude. Since there were more pilgrims there, it became a bigger tragedy. But smaller magnitude similar events are actually quite common (in Uttarakhand and all Himalayan hill states).
The area above 2,200 to 2,500 metre elevation should be organised very thoughtfully because such events are very common and natural in these zones.
After the Chamoli disaster in 2021, what changes have been made in infrastructure planning and early warning systems?
A lot has been done by the government, and a lot still needs to be done, but perhaps its impact is yet to reach the ground level effectively. The 2021 event was significant, and many similar disasters occurred -- small dams were destroyed completely.
However, the Srinagar dam downstream prevented this tragedy from cascading further downstream. Good dams can absorb such events.
Similarly, whatever debris material comes from this current disaster will eventually settle in Tehri Dam. The Tehri Dam will absorb whatever reaches there -- sediment will be transported and settle there, though large boulders won't travel that far.
But this affects the stability and operational life of these mega dams, which needs constant monitoring and assessment.
From current reports and satellite imagery, what's your assessment of the damage?
From all the news we're seeing and satellite pictures we're analysing, it appears that almost half of the settlement area (in Dharali in Uttarkashi) has been destroyed -- it's almost completely flattened.
Regarding casualty numbers, news reports are varying -- some saying 5, others saying 50. The clear picture will emerge in the coming days. The army is currently deployed there for rescue and assessment operations.
It's not appropriate for researchers like us to go there right now because the experts who need to fix immediate problems -- rescue operations, clearing debris, establishing communication -- need to work first.
Once the army completes its immediate response work, detailed scientific assessment can begin.
- Part 2 of the Interview: 'The Problem With Quick Fixes Is...'
Photographs curated by Manisha Kotian/Rediff
© 2025