Coffee & Geography Podcast: Espresso & Geography #11 The Nepal floods with Alistair Hamill (livestream special)
Espresso & Geography co-host Alistair Hamill and I wanted to address the disaster in Nepal and Tibet that is still unfolding. Both of us felt that there was a useful geographical conversation to be had. That also meant being very clear about the limits of what we knew.
When news first broke on 26 August 2026, the initial explanation centred on an earthquake. A seismic signal had been detected around the time the flood began, and in a tectonically active part of the Himalayas an earthquake was an understandable first possibility.
The developing evidence has since shifted the explanation. The US Geological Survey’s preliminary assessment indicates that a rapid slope failure involving a glacier generated seismic energy equivalent to a magnitude 5.2 earthquake. In other words, the seismic signal appears to have been produced by the collapse, rather than an earthquake triggering the collapse. The debris flow and flooding then travelled approximately 100 kilometres through river valleys and populated areas downstream. The USGS continues to describe this as an active investigation, including uncertainty over whether the initial event should ultimately be classified as a landslide incorporating glacier ice or a glacial collapse.
That distinction is important for understanding the process, but it does not reduce the human tragedy. Flooding, mud and debris destroyed homes, roads, bridges and vital infrastructure, while communications disruption and damaged access routes made it difficult for emergency workers to establish the full scale of the disaster or reach affected communities. Rescue and recovery work was continuing as we recorded.
Trying to understand an unfolding event
One of the difficulties of discussing a current disaster is that early information changes quickly. Measurements are revised, casualty figures rise, explanations are tested, and footage circulates faster than it can be verified.
Alistair spoke about initially wondering whether some of the extraordinary footage was real. That is now a necessary response rather than cynicism. AFP has already documented unrelated footage of a glacier collapse in Patagonia being falsely presented online as footage of the Nepal-Tibet disaster.
Some of the genuine footage was also deeply disturbing. We chose not to show it during the livestream. The physical scale of the flood was relevant to the explanation, but the people visible in those recordings were not teaching props. They were experiencing an unfolding emergency.
That created one of the central tensions in our discussion. Images and personal testimony can communicate the reality of a disaster in ways that statistics cannot, but repeated viewing can turn suffering into spectacle. The educational question is not simply whether we show an image. We need to ask why we are showing it, what learners are being invited to notice, whether the people depicted have any agency in the way their experience is being used, and what action or understanding follows.
What appears to have happened
The Nepal disaster was initially described in some reporting as a glacial lake outburst flood. However, the emerging evidence points towards a large mass of ice and rock detaching from a high mountain slope, falling into the valley and developing into a highly mobile debris flow and flood. Satellite imagery and the nature of the seismic signal have been central to this developing explanation.
This was not simply a wall of water. Ice, fragmented rock, sediment, meltwater and material collected from the valley combined into a much denser flow. The extreme relief of the Himalayas converted gravitational potential energy into rapid movement downslope, while narrow valleys channelled the flow downstream.
Alistair explained this in detail during the episode. The substantial vertical drop, the pulverising of ice and rock, the presence of water, and the entrainment of additional material all helped create an exceptionally destructive flow. Monitoring infrastructure was itself damaged or destroyed, reducing the ability to relay information downstream.
As geographers, we often separate processes into categories because that helps us teach and analyse them. Landslide. Avalanche. Debris flow. Flood. Glacial hazard. In reality, mountain disasters can cascade from one process into another. This event appears to have moved rapidly across those boundaries.
What can we say about climate change?
We were cautious about this during the livestream and remain cautious now. A formal attribution study cannot be completed within the first days of an event.
The immediate physical cause appears to have been a major slope failure involving ice and rock. Establishing exactly why that part of the mountain failed will require further analysis of weather, glacier structure, bedrock, slope geometry, meltwater, permafrost and previous conditions.
However, caution does not require us to pretend that the disaster happened outside a warming climate.
Zeke Hausfather’s early analysis for The Climate Brink makes the distinction clearly: formal attribution is premature, but the collapse occurred on a rapidly warming mountain in a valley losing glacier ice. Supporting climate data show long-term warming in the area and provide context for the conditions in which this event occurred.
Long-term human-caused warming is destabilising glaciers and permafrost across the Himalayas, creating conditions in which ice and rock failures can become more likely, although this broader relationship is not the same as proving that climate change caused this individual collapse.
During the episode I used the familiar “loaded dice” analogy. Climate change does not necessarily create an entirely new hazard. Instead, it can alter the background conditions, making some outcomes more likely or more severe when the dice are rolled.
The important thing is to communicate both parts of that sentence. We should not overclaim what current evidence can establish, but nor should uncertainty be misrepresented as evidence that climate change is irrelevant.
A disaster crossing political borders
The flood also demonstrated that physical processes do not stop at political boundaries.
The river system crosses the Nepal-Tibet border, while monitoring, warning, emergency response and information-sharing remain organised through political institutions. The physical hazard was transboundary, but the ability to observe it and communicate risk depended on cooperation between different authorities.
A peer-reviewed paper on climate-induced loss and damage in Nepal was published in Frontiers in Climate on 14 July 2026, only weeks before this disaster. The paper did not predict this individual event. It did, however, describe Nepal’s vulnerability to climate-related hazards including flooding, landslides, glacier retreat and glacial lake outburst floods. It also identified fragmented institutional responsibilities, limitations in the availability and quality of data, and the need for stronger inter-agency coordination and physics-based attribution capability.
That research is a useful reminder that the warnings surrounding disaster risk are often already present. The difficulty lies in turning knowledge into monitoring, infrastructure, communication, finance, policy and action before the disaster occurs.
Why do some disasters remain visible?
Our conversation then widened into the way disasters are represented.
Some events become part of global cultural memory. The 2004 Indian Ocean tsunami is one example. The 2011 Tōhoku earthquake and tsunami is another. Both were enormous disasters, but their visibility was also shaped by the volume and immediacy of recorded imagery.
Alistair went further back to the 1985 Nevado del Ruiz eruption and the destruction of Armero in Colombia. The disaster became associated internationally with the ordeal of one child, whose final days were recorded by news cameras. Alistair remembered following those reports as a young person beginning to fall in love with geography.
That story raises difficult questions. A person’s experience can humanise a disaster which might otherwise become a list of statistics, but the repeated circulation of suffering can also become intrusive or voyeuristic.
The problem is not easily solved by refusing to teach current disasters. Young people will see footage on social media whether teachers discuss it or not. Avoiding the subject can leave them alone with misinformation, graphic material and algorithmically amplified speculation.
The response should be careful geographical education: verification, context, process, empathy, agency and space to ask questions.
The disaster that many people did not see
Alistair used the 2023 flooding in Derna, Libya, as an example of a disaster which did not remain in public consciousness to the same extent.
Storm Daniel brought extreme rainfall to eastern Libya, while the failure of two dams upstream released a devastating flood through the city. Thousands of people were killed, many more were reported missing, and tens of thousands were displaced. The exact figures were revised repeatedly after the disaster, another reason to be careful when presenting early estimates as settled facts.
The physical geography cannot be separated from Libya’s political history. Conflict, divided governance, infrastructure neglect and failures of maintenance helped turn an extreme weather event into a far larger catastrophe. Warnings existed, but the capacity to translate those warnings into effective protection and evacuation had been weakened.
This is what geographers mean when we say there is no such thing as a natural disaster. The storm was a hazardous atmospheric process. The rainfall was real. The floodwater was physical. However, the scale and distribution of harm were shaped by decisions about infrastructure, maintenance, planning, communication and governance.
Teaching the human geography of hazards
Alistair and I both shared examples of how we have tried to teach disasters through personal testimony.
I spoke about teaching the 2010 Haiti earthquake through the reported experiences of Mexican rescue worker Héctor Méndez and Los Topos. I used a careful form of teacher-in-role, answering pupils’ questions using information that had been attributed to Méndez in interviews and reporting rather than inventing thoughts or experiences for him.
The intention was to move beyond the standard “causes, effects and responses” structure without abandoning geographical knowledge. Pupils still learned about tectonic processes, vulnerability, infrastructure, colonial history, governance and emergency response, but they encountered those issues through a named person’s experience.
I also shared how I talked about a wildfire near Bozeman, Montana. The fire received far less international attention than the larger fires elsewhere in western North America, but it came close to a house used by my family. Through people I knew in the area, I was able to seek permission to use photographs, testimony and local accounts.
That personal connection changed the way I understood and taught the event. It was not more important than other fires because it affected somewhere connected to me, but it reminded me that every apparently “small” disaster is central to somebody’s life.
Alistair and I have both written articles for the journal Teaching Geography focussing on how to teach disasters such as these. You can download these for free at the end of this post.
Unequal outcomes in Guatemala
Alistair closed with his teaching about the June 2018 eruption of Volcán de Fuego in Guatemala.
The eruption affected communities including San Miguel Los Lotes and the La Reunión resort. The eruption produced pyroclastic flows and lahars, with almost 200 deaths recorded officially and many more people reported missing.
Alistair’s inquiry compares different communities exposed to the same volcanic activity. Residents and visitors at the resort had greater mobility and access to evacuation, while people in the informal settlement at Los Lotes faced very different constraints.
The physical hazard did not produce an equal outcome because the people exposed to it did not have equal resources, mobility, political power or protection.
Alistair ends the case study by telling the story of a disaster-management worker who drove into the hazardous area to warn residents and lost his life during the attempt. Each year, Alistair speaks the person’s name in a classroom in County Armagh and gives pupils time to reflect.
He finished our episode by explaining why:
“We remember the people, not, I hope, in a voyeuristic way, not in an exploitative way, but in a way that honours their suffering, honours their bravery, honours their sacrifice, and honours their memory.”
That felt like the right place to stop.
What happens after the media move on?
At the time of writing, emergency operations in Nepal and Tibet are continuing. Access remains difficult in some locations, communities face further risks, and casualty figures continue to change. The IFRC has launched an emergency appeal supporting the Nepal Red Cross Society’s delivery of shelter, clean water, first aid and emergency supplies. UNICEF has also issued a flash appeal for its response in Nepal.
Media attention will move elsewhere, as it always does. The people affected will continue living with bereavement, displacement, damaged infrastructure, lost livelihoods, interrupted education and a landscape altered within minutes.
If teachers return to this disaster, I hope the lesson is not built only around dramatic footage or an examination-friendly list of primary and secondary effects. There is valuable physical geography here, but there are also questions of political borders, early warning, inequality, climate justice, media representation, consent, recovery and whose stories are remembered.
Geography helps us understand what happened. It should also help us remember that it happened to people.
