Written by A.N Mohammed
Every year, the waters rise – floods come. Every year, Assam suffers; lives are uprooted, the economy is crushed. Behind every flood statistic is a family rebuilding from loss. Assam’s resilience is inspiring, but resilience alone cannot be the solution. The time for lasting change is now.
The scale of the crisis
Few states face flooding and erosion as severe, frequent, and prolonged as Assam does. The Rashtriya Barh Ayog estimates that 31,050 sq km- 39.58 percent of the state is flood-prone. Riverbank erosion has steadily consumed Assam for more than six decades, taking 4,270 sq km, or 7.4 percent of the state, since 1950.
Majuli, the world’s largest river island, illustrates the toll: it shrank from 787.9 sq km in 1915 to 508.2 sq km in 2005, losing 35.5 percent of its area at an average rate of 3.1 sq km a year. Government estimates place annual flood damage at about Rs 3,500 crore, but the true cost, including human suffering and wider economic disruption, is widely believed to exceed Rs 10,000 crore each year.
A flood unlike any before
Assam has just endured an unprecedented flood. The Dikhow surge reached Sivasagar district on 19 July 2026 and, by 21-22 July, became the region’s worst flooding on record. It submerged large parts of Sivasagar, Charaideo and Jorhat districts and affected nine lakh people at its peak.
While the flood situation has shown signs of improvement in certain areas, the crisis remains a matter of serious concern. As of 6 August 2026, more than 1.68 lakh people continue to be affected, and the confirmed death toll has risen to 97.
The floods have also resulted in substantial losses of livestock and poultry, while extensive tracts of agricultural land remain buried under thick deposits of silt, severely impairing their agricultural productivity.
The flood’s speed was starkest at Nepalikhuti, a self-reliant dairy village in the Nazira sub-division of Sivasagar district. On the morning of 19 July, the Dikhow river, swollen by heavy rain in the Nagaland hills, became a wall of water. It breached its embankments, swept away nearly 200 homes in under an hour, and submerged the village within two hours, between 11 a.m. and 1 p.m. The devastation there raises the question that must guide any serious reckoning with this disaster: what turned heavy rain into such a sudden and destructive surge?
What triggered it
Extreme rainfall over Nagaland and upper Assam between 18 and 20 July triggered the disaster. During that period, Mokokchung recorded 238.9 mm of rain, Sonari 197.5 mm, Jhanji 181.7 mm and Morigaon 122 mm, while Charaideo received 436 percent more rain than normal and Sivasagar 134 percent more. In the hills near Nagaland’s Mon district, heavy rain also caused landslides due to deforestation.
Nagaland has lost roughly 280,000 hectares of tree cover between 2001 and 2025, a 21 percent decline since 2000, according to Global Forest Watch, though only about a tenth of that is traced to outright deforestation. Shifting cultivation, or jhum, drives the bulk of the loss, accounting for some 220,000 hectares.
Within the deforestation share itself, logging (26,000 hectares) and permanent agriculture (24,000 hectares) account for most of the damage, followed by settlements and infrastructure (2,900 hectares), while natural disturbances (1,300 hectares), hard commodities such as coal mining (670 hectares), and wildfire (110 hectares) account for relatively minor shares of the total loss.
One possibility is that debris from a landslide, or rejects from coal mining, blocked a stream, formed a temporary lake and then burst suddenly downstream, adding a powerful surge to the rain-driven flood. This is known as a landslide lake outburst flood, or LLOF.
The India Meteorological Department has denied that a true cloudburst of more than 100 mm of rain in an hour occurred. Even so, experts note that rainfall alone cannot explain destruction on this scale. Weak embankments narrowed river channels, and shrinking wetlands helped turn a severe flood into a catastrophe.
The deeper, structural causes
This year’s rain exposed decades of structural decline. Since the 1950 earthquake, sediment has steadily built up in the Brahmaputra and upper Assam’s tributaries, reducing their capacity so even moderate flows now spill over. Settlements, roads and commerce have encroached on floodplains, while the loss of beels, wetlands and low-lying areas has weakened the landscape’s ability to absorb flood peaks.
Deforestation, landslides and hill-cutting have further increased runoff and silt. As a result, Assam has faced floods of varying severity almost every year for more than seven decades, with climate change now adding short, intense bursts of rain that can turn tributary basins into flash-flood zones with little warning.
Assam’s history offers a sobering reminder of the risks posed by landslide-dammed rivers. The 1950 earthquake triggered massive landslides on both banks of the Subansiri near the Sippu Nalah confluence, about nine kilometres upstream of Gerukamukh in Dhemaji district, depositing an estimated 19 million cubic metres of debris on the right bank and 10 million on the left. The blockage dried the downstream riverbed for three days before giving way, sending a violent surge through downstream Assam and killing 532 people.
A smaller but comparable mechanism was seen at Aberfan in Wales in 1966, when a colliery-waste tip collapsed after days of rain, burying a village and its school and killing 144 people. Whether created by a landslide or by mining debris, loose material that blocks a channel remains a serious hazard until it is safely drained and stabilised.
Embankments: Necessary, but not the permanent answer
Assam’s flood-control strategy began after the devastating floods of 1954, which led the Government of India to frame its first National Policy on Floods. The state started its own flood-management works during the Second Five-Year Plan (1956-61), followed by a detailed flood-control blueprint after the 1987 National Water Policy. Over time, this effort created 4,474 km of earthen embankments, 911 anti-erosion and town-protection works, and 875 km of drainage schemes.
Embankments increase a river’s carrying capacity, but most of Assam’s 4,474 km network is now more than fifty years old. Built as an immediate defence against a 25-year-return-period flood, it was never meant to be a permanent solution. Many stretches are now fragile, and breaches occur almost every year.
Strengthening the existing network with geo-bags, rock and concrete spurs, pile foundations, and modern construction methods is therefore more urgent than building new embankments. Dredging the Brahmaputra and its tributaries may briefly improve flow, but without upstream silt-trapping measures, the channels soon fill again. Neither embankments nor dredging, on their own, address the roots of Assam’s flood problem.
The case for large reservoirs
China’s Huang He, long known as the ‘Sorrow of China’, is often cited as an example of how dams can help moderate silt and flood flows while also supporting power generation and irrigation. Many water-resource experts in India argue that Assam, too, needs large reservoirs capable of temporarily absorbing flood peaks. The contributions of the Siang, Subansiri, Dibang, and Lohit rivers to the Brahmaputra are approximately 25, 10, 8, and 7 percents respectively. Since these tributaries collectively contribute half of the Brahmaputra’s flow, controlling floods in these rivers could potentially mitigate flooding in Assam by 50 percent.
This idea is not new. In 1955, the Assam government’s Brahmaputra Flood Control Commission began surveying multipurpose dam sites on the Siang and Subansiri rivers. From 1982, the Brahmaputra Board conducted scientific studies and proposed projects on key Brahmaputra tributaries.
By 1983, a 295-metre, 20,000 MW project on the Siang near Pasighat and a 257-metre, 4,800 MW project on the Subansiri at Gerukamukh had received technical approval from the Government of India. Together, they were expected to lower Brahmaputra flood levels at Pandu by about one metre. Both projects, however, were set aside after Arunachal Pradesh raised concerns over the extent of submergence, shifting planning toward smaller and medium multipurpose and hydropower projects on individual tributaries.
In 2016, the Centre proposed 50 storage ponds along the Brahmaputra between Sadiya and Majuli to absorb about 20,000 cumecs of excess flow the difference between the river’s safe carrying capacity of 40,000 cumecs and a 24-hour flood peak of roughly 60,000 cumecs, or about 1,728 million cubic metres of water. Each pond would need to hold around 35 MCM, requiring a 10-metre-deep basin spread over about 350 hectares.
With intake and outlet canals, gates, pumps and silt-removal systems, the footprint would rise to nearly 400 hectares per pond about 20,000 hectares for all fifty. Given the land acquisition, cost, and uncertain benefits involved, the scheme’s practicality remains open to serious question.
A persistent misconception
In Assam, floods on almost any river are often blamed on nearby hydropower projects. But the reservoirs at Kurichu in Bhutan, Ranganadi in Arunachal Pradesh, Kopili in Assam and Doyang in Nagaland are too small to store major flood volumes. When inflows rise, they must pass the water downstream, keeping the river’s natural flow largely unchanged.
During a flood wave, operators open spillway gates wider to maintain the reservoir at Full Reservoir Level and prevent overtopping – a dam-safety measure, not the cause of flooding. Unlike the Dikhow, these projects also issue real-time warnings to downstream areas. True flood moderation requires high dams and large reservoirs, which these smaller projects were never designed to provide.
The Thatte-Reddy Committee, reviewing the Subansiri Lower Project, put the contradiction plainly: agitating against a dam while demanding flood control is self-defeating, since there can be no flood control without a dam, and it is hard to see the case against dams built to state-of-the-art standards.
The under-construction 278-metre, 2,880 MW Dibang Multipurpose Project has reserved 41 metres of its reservoir purely for flood cushioning, designed to contain even Dibang’s hundred-year flood fully. The partially commissioned, 116-metre, 2,000 MW Subansiri Lower Project already reserves 15 metres, about 442 million cubic metres, for flood control, enough to hold back Subansiri’s ten-year flood for 24 hours.
Once the proposed 1,720 MW Middle Subansiri and 1,605 MW Upper Subansiri projects are built and operated jointly with it, the Subansiri’s hundred-year flood could be fully controlled. The value of even partial capacity was shown on 13 July 2026, when an early alert of a midnight surge let authorities warn downstream villages by siren in time, cutting the damage from a flood that lasted four to five hours the next afternoon. NHPC’s early-warning systems at Daporijo and Tamen, which gather real-time flood data roughly six hours ahead, have similarly moderated most of the smaller, short-duration floods on the Subansiri over the past two years.
The way forward
This year’s floods were intensified by a combination of extreme rainfall, a suspected landslide dam in the upstream catchment, and rising water levels across the Brahmaputra and its tributaries. Yet, this disaster cannot be dismissed as a purely natural calamity. It is also a stark reminder of longstanding shortcomings in flood management, planning, and infrastructure.
Assam’s continued dependence on ageing embankments has left communities, farmland, and critical infrastructure highly vulnerable, underscoring the urgent need for their redesign and reconstruction using modern engineering and technology. Equally important is a scientific assessment of the catchments of all river basins, followed by systematic treatment through afforestation and targeted engineering interventions in erosion- and landslide-prone areas. All riverbed mining activities should be confined to the central one-third of the riverbed, while the one-third sections adjacent to both banks should remain undisturbed.
The state now requires a comprehensive, technology-driven flood management strategy for its tributaries, supported by joint flood forecasting by the Central Water Commission and the India Meteorological Department for every river in the Northeast. Satellite-based surveillance should be deployed to detect the early formation of landslide dams in hill catchments, while round-the-clock basin monitoring and robust early-warning systems, on the model of the Subansiri system, should be extended to all vulnerable regions.
Large flood-cushion reservoirs, such as the Dibang and Subansiri Lower projects, can play a crucial role in moderating flood peaks and breaking Assam’s recurring cycle of flood damage, relief, and rebuilding.
Above all, the Brahmaputra Board’s master plan must be urgently revised to account for climate change realities and be backed by adequate funding, statutory authority, and effective inter-state coordination. The time for piecemeal responses is over. Governments at the Centre and in the Northeast must act now with urgency and vision, transforming flood management from a seasonal relief exercise into a long-term mission of resilience and adaptation. Without decisive action today, Assam will continue to pay an avoidable human and economic price with every monsoon.
A. N. Mohammed is a consultant in hydro project development