On August 26, 2026, a devastating flood struck the Nepal-Tibet border region, killing more than 1,200 people and leaving over 5,000 local and foreign nationals missing as of September 3, 2026.
However, this is not the first time that unpredictable natural disasters have caused extensive harm to people, the natural environment and biodiversity and economy, indicating the rising frequency of such events.
In the current scenario, states can no longer treat disaster preparedness as an exercise in matching response capacity to historical hazard patterns. They need systems versatile enough to meet hazards that have no local precedent at all.
High-velocity mountain floods carrying large volumes of debris are among the most destructive of these hazards. They combine the speed of a flash flood with the crushing force of a landslide, and within hours they can dismantle a region's economic and social fabric alongside the direct human cost they exact.
Assam has experienced two severe mountain-flood disasters within four years. Heavy rainfall during May 11-18, 2022 triggered floods and landslides in Dima Hasao, with satellite mapping identifying 5,178 landslides. Three people died, nearly 57,000 residents were displaced, over 200 homes were destroyed, and road and rail infrastructure worth several hundred crore rupees was damaged.
A 2025 study found that residents were also badly affected by the subsidence of land, which reduced habitable and cultivable land and left many concrete houses damaged or abandoned. Residents attributed this to earlier mining, which had left the ground filled with loose sand and unable to withstand heavy rain.
A similar pattern emerged in the July 2026 floods in Upper Assam, on a larger scale and further downstream. Independent accounts linked the disaster to deforestation and mining in the hills around Naginimora, to unregulated sand and stone extraction from the Dikhow riverbed, and to rathole and opencast coal mining across five Nagaland districts bordering Assam. In both regions, extraction stripped the hills of the vegetation and soil structure that held them together. It was this weakened ground, rather than rainfall alone, that converted ordinary monsoon rain into a mountain flood disaster.
The 2026 floods also exposed the limits of Assam's disaster management framework, which assumes slow-onset, Brahmaputra-driven flooding in familiar low-lying districts. Rapidly rising rivers sent flash floods into Sivasagar, Charaideo and Jorhat, which recorded their worst inundation in over six decades. More than 60 people died, and some 700,000 were affected. Communities had almost no warning and no precedent to draw upon.
According to a stakeholder at the District Disaster Management Authority, Dima Hasao, the equipment, funding, and training remain predominantly flood-focused, with mountain-specific provisions concentrated in Uttarakhand. The recurrence of such events in two locations within five years indicates that they are no longer anomalies but a standing feature of the state's hazard profile, to which institutional priorities have yet to adapt.
This is not a gap that exists only in the perception of one district's stakeholders. It shows up consistently across official strategy documents, force-level admissions, and the state's own recent procurement choices.
An editorial written after the 2022 Dima Hasao landslides pointed out that NDMA's own landslide strategy already lists the northeastern states as risky. Even so, the strategy was not visible on the ground.
After the 2021 Uttarakhand disaster, new investment went to the west and north. NDRF's plan for permanent mountain rescue teams named only Jammu and Kashmir, Uttarakhand, Himachal Pradesh and Sikkim.
In January 2026, ASDMA's main equipment purchase was flood-rescue boats, and no equipment for mountain- or landslide-rescue was announced.
What Assam needs to do next is not complicated to describe. Doing it will be the hard part. Assam does not need a flood response system that is occasionally asked to handle a landslide, or a landslide-aware policy paper that is occasionally funded. It needs to stop building its disaster architecture around the most familiar hazard and start building it around the full range of hazards the state actually faces.
This means SDRF units trained and equipped for hill terrain and debris flow, not only for river rescue. It means mitigation funds that reach slope stabilisation and catchment restoration in the same proportion they reach embankments. It means treating upstream ecological degradation, whether from mining, deforestation or unplanned construction as a disaster risk factor in its own right, not an afterthought raised only once the damage is done.
Two unpredicted mountain floods in five years are not an anomaly Assam can wait out. They are early evidence of a hazard profile the state has not yet built for. What the moment calls for is a dynamic disaster response system, one versatile enough to meet whichever emergency the terrain produces next, rather than the one it produced last.
Mrinal Saikia is assistant professor, department of economics, Cotton University, Guwahati, Assam, India. Views expressed are the author’s own and don’t necessarily reflect those of Down To Earth.