Sep 05: A water scientist at the University of Rhode Island, Soni Pradhanang is balancing the professional and the personal, watching her home country of Nepal make news after the recent glacial collapse and flooding there.
As a hydrologist with expertise in environmental geology, flooding, and hazard communication, the events in her home nation are the subject she has made her career in since she left Nepal in 2001. Pradhanang, in URI’s College of the Environment and Life Sciences, conducts research in runoff, flood, drought, and water quality modeling, and has served as a point of contact in URI’s prior work with Nepal.
In 2022, during her Fulbright tenure, she served as an instructor for a hydrological modeling training program in Kathmandu, part of a project on
“Modeling Hydrological Response to Changes in Climate and Glacier in a Data-Scarce Glacierized Region in Nepal Himalaya.” The training covered core concepts in hydrology, climate change, and cold-region hydrology, providing early-career professionals with the tools to study these issues. “This tragedy highlights the urgent relevance of this research,” she said.
Raised in Kathmandu, approximately 40 miles aerial distance from the affected area in Rasuwa/Langtang National Park, she shares insight into recent events, while keeping tabs on family members and loved ones there.
“For me, this is not an abstract scientific event,” Pradhanang said. “It is a human tragedy affecting communities, families, livelihoods, and landscapes in my home country.”
What were the contributing factors for the recent events in Nepal and has anything like this happened there before?
Based on the best available satellite imagery, seismic observations, and preliminary field interpretation, the disaster in northern Nepal appears to have begun with a rapid failure of a glacierized mountain slope in Langtang National Park near the Nepal–China border. A very large mass of ice, rock, and debris descended into the valley, likely obstructing the Lhende Khola River temporarily. When that unstable blockage failed or was overtopped, it released a powerful debris-laden flood that traveled downstream through the Lhende and Trishuli river valleys.
It is important to be careful about the cause. The collapse itself generated seismic energy comparable to a magnitude 5.2 event; it should not automatically be described as an earthquake-triggered disaster. Scientists worldwide are still determining the relative roles of steep terrain, fractured bedrock, glacier geometry, meltwater, rainfall or snowmelt conditions, and thawing permafrost. Climate warming is a highly relevant background factor because it causes glacier retreat and can thaw the frozen ground that helps stabilize high-mountain rock and ice. But attributing this particular collapse to one single factor would be premature.
Nepal has experienced destructive glacier- and landslide-related floods before, although every event has its own sequence of triggers. Nepal has recorded 26 glacial lake outburst floods since 1977, including events originating both within Nepal and beyond its borders. The Langtang region also has a history of serious mountain hazards, including earthquake-triggered avalanches in 2015 and recent floods and ice-rock avalanches in nearby valleys. This event is especially notable because it appears to involve a rapid ice-and-rock avalanche and a temporary river-dam failure, rather than a conventional moraine-dammed glacial lake suddenly bursting. A similar devastating event occurred in India in 2021, involving a rock-ice avalanche. More recently, in 2024, a cascading glacial lake outburst flood event in Nepal’s Thame Valley, triggered by a rock avalanche and a chain reaction of lake breaches, caused widespread destruction. These past disasters highlight the growing danger.
Is there any technology that could provide a warning for an event like this?
No system can predict the exact moment when every slope or glacier will fail, or when a rock-ice avalanche will occur; that would defeat essentially every operational flood warning system in the world. However, technology can provide critical warning—especially once a collapse, blockage, or sudden rise in river level begins.
A strong system would combine satellite monitoring with ground-based instruments. Satellites can detect changes in glacier extent, new or expanding lakes, ground deformation, slope movement, and post-event flood extent. On the ground, river-stage and rainfall gauges, time-lapse cameras, geophones or seismic sensors, and automated alarms can detect rapidly developing hazards. Warnings then need to be distributed immediately via sirens, cell broadcasts, radio, local responders, and community networks.
The limitation is time. In narrow, steep valleys, a debris flood may reach downstream settlements in minutes. Therefore, the system must be automated, resilient to damage, powered and connected with backups, and linked to practiced evacuation plans. Monitoring alone does not save lives unless people receive, understand, trust, and can act on the warning.
What can be done to lessen impact in events like this in the future?
We cannot eliminate all high-mountain hazards, particularly in a country with steep terrain, active geology, and rapidly changing glaciers. But we can substantially reduce loss of life and damage. While predicting these sudden events is incredibly difficult, especially when no immediate rainfall precedes them to trigger traditional flood warnings, several measures can be taken to mitigate the risks and protect vulnerable communities.
The most important actions are to identify hazardous slopes, glaciers, and glacial lakes; map the downstream pathways that floodwaters and debris would follow; install robust monitoring and warning systems; and ensure that communities, schools, hydropower operators, road agencies, and border facilities know exactly how to respond to an alert.
Risk reduction should include four inter-connected elements:
- Regular satellite and field monitoring of glacierized slopes, glacial lakes, landslide-prone areas, and thawing permafrost, if any.
- Multi-hazard early-warning systems using river-level sensors, cameras, weather stations, seismic instruments, satellite data, and redundant communications.
- Community preparedness: alarms, evacuation routes, safe assembly areas, drills, and warning messages that are understandable and accessible to everyone.
- Safer planning and infrastructure. Critical facilities such as bridges, hydropower stations, roads, schools, and settlements should, when alternatives exist, be avoided in the highest-hazard flood corridors. Where exposure already exists, structures and evacuation systems should be designed for rapid-onset debris floods, not only ordinary monsoon flooding.
For selected high-risk glacial lakes, direct interventions can also help—for example, carefully lowering lake levels, reinforcing or modifying unstable moraine dams, and coupling those interventions with locally managed warning systems. Nepal has already implemented lake-lowering and community early-warning measures at Imja Tsho in the Everest region, demonstrating that targeted mitigation is possible.
Was the glacier that triggered the event considered vulnerable and how do researchers assess glaciers?
At this early stage, it is more accurate to say that the glacierized slope was demonstrably hazardous than to imply that the precise failure was fully predictable. Researchers assess vulnerability by examining satellite imagery and aerial data, glacier retreat and surface changes, slope angles, fractures and crevasses, rock and ice volumes, evidence of permafrost thaw, nearby lakes, drainage pathways, and the likely consequences for people and infrastructure downstream.
Researchers assess the vulnerability of glaciers and glacial lakes, using satellite imagery to map their size, evolution, and growth over time. This helps identify which lakes are expanding rapidly and are therefore potentially dangerous.
Pratik Singh Thakuri, a Ph.D. student in my lab also from Nepal, has been actively acquiring and analyzing satellite imagery to better understand both the physical processes that drove this event and the extent of resulting damages. Trained as a water resources engineer, he focuses his research on modeling floods and other extreme hydrologic events. His work provides an important connection between remote-sensing observations of this disaster and the broader effort to understand, model, and reduce flood risks.
Pratik’s comparison of satellite-based images, from Aug. 24 and Aug. 26, 2026, look at the two scenes over the same geographic area, comparing the glaciers, surrounding slopes, avalanche pathways, valley floods, and downstream river corridors. By comparing different time periods, scientists can piece together the development of an event and ascertain whether there were any visual changes in the glacier or adjacent region prior to the event. Although this does not identify the cause of the event, it represents an important starting point for analyzing it.
When possible, teams conduct on-the-ground surveys to study glacial conditions, measure lake depth, assess the stability of moraine dams, and gather data for computer models. Scientists integrate data on lake size, surrounding topography, and downstream population and infrastructure to assess risk. A lake might be deemed “high risk” if a potential outburst could impact a populated area. We use computer models to simulate what would happen if a lake were to breach, predicting the path, speed, and volume of a flood wave.
The key question is not only ‘Could a glacier or lake fail?’ but also ‘Who and what lies in the potential flow path?’
There are other sites of concern in Nepal and across the Hindu Kush Himalaya. A 2020 assessment, by the International Centre for Integrated Mountain Development / United Nations Development Programme, identified 47 potentially dangerous glacial lakes across the Koshi, Gandaki, and Karnali basins spanning Nepal, Tibet, and India—21 in Nepal, 25 in Tibet, and one in India. These inventories are vital but need to be updated because glaciers, lakes, slopes, and downstream exposure are changing rapidly.
Could this kind of an event happen elsewhere and are there other areas that might be vulnerable or at-risk?
Similar cascading hazards can occur wherever steep, glacierized mountains contain unstable ice, thawing permafrost, fractured rock, glacial lakes, and communities or infrastructure in narrow downstream valleys. The Hindu Kush Himalaya is a major concern, but the underlying processes also occur in the Andes, Alaska, western Canada, the European Alps, Iceland, and parts of Central Asia. In a paper I co-authored in 2023, we discussed cascading hazards becoming more prevalent in the central Himalayas and how the need for an integrated modeling framework means more reliable prediction, with early warning systems essential for reducing impacts.
The specific hazard may differ by place. In some valleys, the major concern is a glacial-lake outburst flood; elsewhere, it may be an ice avalanche, a rock-and-ice avalanche, a landslide-dammed lake, a rain-on-snow flood, or a compound event involving several of these processes. Warming does not make every individual collapse directly attributable to climate change, but it is altering the conditions that affect glacier and permafrost stability in many mountain regions.