Assam and the 1% Flood
The most flood-exposed state in the atlas — where nearly half of every building stands in the Brahmaputra's path
There is a number the Government of India has quoted for half a century: 39.58% of Assam is flood-prone — about 31.05 lakh hectares, and roughly four times the national average. It comes from the Rashtriya Barh Ayog, the national flood commission, and it has shaped every embankment, every relief budget and every argument about the Brahmaputra since.
This atlas was built from an entirely different direction. It does not start from land classification or historical flood records. It models a 1% annual-chance flood on a ~30 m elevation model, then counts how many actual building footprints fall inside the resulting water. In Assam, that method returns 43.7%.
Two independent approaches — one official and area-based, one modeled and building-based — landing within a few points of each other is the strongest corroboration in this atlas. Assam is not merely a flood-prone state. It is the most flood-exposed region the India Flood Atlas has mapped.
The scale
- Of Assam’s 12,997,670 building footprints, 5,675,946 — 43.7% — sit in the modeled 1% annual-chance floodplain. That is close to one building in every two. No other region in the atlas approaches it; the national figure is around 20%.
- The mean depth over exposed buildings is 1.45 m — and unlike the hill states around it, this number is trustworthy. The Brahmaputra valley is flat, which is exactly the terrain a 30 m DEM handles well. Roughly waist-to-chest deep across millions of buildings.
- 45,859 km of road — 17.9% of the network — lies in the floodplain.
- 2,530 km of railway — 38.9% of the network — is exposed. This is the highest rail exposure in the atlas by a wide margin, and it is not an accident: the trunk line through the Brahmaputra valley is the only viable corridor, and the valley is the floodplain. When Assam’s rail link goes down, so does the surface connection to the whole of the Northeast.
Note the asymmetry between depth and extent. At 1.45 m, Assam’s flooding is not exceptionally deep — Kerala and Odisha model deeper. What makes Assam singular is how much of the state the water reaches, and how often.
Where it concentrates
The pattern is the river itself. Exposure is overwhelming along the Brahmaputra’s braided lower and middle course, and it peaks where the river is widest and the land flattest.
| District | % of buildings exposed | Exposed buildings | Mean depth |
|---|---|---|---|
| South Salmara Mankachar | 89.6% | 168,065 | 2.01 m |
| Majuli | 88.6% | 57,817 | 1.74 m |
| Barpeta | 81.3% | 537,432 | 1.93 m |
| Dhubri | 78.6% | 407,962 | 1.52 m |
| Morigaon | 71.6% | 244,471 | 2.04 m |
| Nagaon | 63.3% | 414,239 | 1.46 m |
| Kamrup | 54.9% | 450,277 | 1.45 m |
| Lakhimpur | 53.3% | 261,639 | 1.09 m |
Four districts sit above 78%. In South Salmara Mankachar and Majuli, nearly nine buildings in ten are in the floodplain — figures that in most states would indicate a modelling error, and here simply describe the geography. Barpeta carries the largest absolute count at 537,432 exposed buildings, more than the entire exposed stock of several Indian states in this atlas.
Majuli deserves its own sentence. It is the world’s largest river island, and it is being consumed: roughly 60% of its original area has already been lost to erosion. An 88.6% exposure figure on an island that is itself shrinking is not a forecast of risk — it is a description of a place already losing the argument with the river.
Why Assam floods
- The Brahmaputra is a sediment machine. It drains the young, rapidly eroding Himalaya and carries one of the highest sediment loads of any river on Earth. That sediment settles in the valley, raising the riverbed, which pushes the same volume of water over a wider area each decade. The river braids across a bed that is rising underneath it.
- The 1950 Assam–Tibet earthquake. At magnitude 8.6, it remains one of the largest continental earthquakes ever recorded. It destabilised whole mountainsides and dumped an extraordinary pulse of debris into the drainage, raising the riverbed further. Assam’s flood regime today is still partly the aftermath of a single day in 1950.
- Extreme, concentrated monsoon rainfall — including runoff from Meghalaya’s Khasi hills, among the wettest places on Earth, which drains into the Barak system and the southern valley.
- A confined valley. The Brahmaputra valley is a narrow trough between the Himalaya and the Meghalaya plateau. Water arriving from both walls has nowhere to spread except across the valley floor, where everyone lives.
- Embankments that have become part of the problem. About 5,000 km of embankments have been built over six decades. Many are ageing, poorly maintained and breach under pressure — and when they hold, they cut the floodplain off from the sediment that would otherwise build it up, while raising water levels against the next embankment downstream. The 2022 Silchar disaster began with a dyke breach at Bethkundi.
- Erosion, which is the quieter catastrophe. Assam has lost about 4.27 lakh hectares — 7.4% of the state — to river erosion since 1950, at roughly 8,000 hectares a year. Flooding displaces people temporarily; erosion removes the land permanently.
- Two separate flood systems. The Brahmaputra valley and the Barak valley in the south are distinct basins with distinct behaviour. Cachar and Silchar flood from the Barak, not the Brahmaputra — which is why a “good year” on one river means nothing for the other.
What this looks like on the ground
2022 is the reference year. Floods and landslides affected 5.4 million people across 32 of Assam’s districts and killed over 200. In Cachar district alone, 1.43 million people were affected. The town of Silchar stayed underwater for 11 days, with water reaching 12 feet in places, after the Barak embankment failed.
The atlas’s district figures show why that scale is structural rather than exceptional. When 43.7% of the building stock is inside the 1% floodplain, a severe monsoon does not need to be unprecedented to displace millions. It only needs to be normal.
Kaziranga National Park floods almost every year as part of its natural cycle — the annual inundation sustains the grasslands the one-horned rhino depends on. But as embankments and highways have hardened the edges, animals fleeing to the Karbi Anglong highlands must now cross a national highway to reach safety, and drownings and vehicle strikes have become a recurring toll. It is a useful reminder that in Assam the flood is not simply a disaster to be eliminated; it is the process that built the valley, now colliding with the infrastructure laid across it.
Reading the numbers honestly
Unlike the hill states in this atlas, Assam’s depth figures are reliable. A ~30 m DEM struggles in near-vertical terrain — that is why Arunachal Pradesh, Nagaland and Sikkim carry explicit depth caveats. The Brahmaputra valley is the opposite case: broad and flat, where the elevation model performs at its best. Both the 43.7% extent and the 1.45 m mean depth can be read at face value.
Two limits worth stating:
- The atlas models riverine flooding for a 1% annual-chance event. It does not model erosion, which in Assam causes permanent loss that no depth map can express. The true damage is worse than these figures show.
- District boundaries follow the 2011-census-derived set (33 districts), so the most recently created districts — Bajali and Tamulpur — are folded into their parent districts rather than shown separately.
A strategic approach
Pillar 1 — Treat erosion as a first-class hazard. Assam loses land permanently every year. Erosion needs its own budget line, its own monitoring, and its own compensation framework, separate from flood relief.
Pillar 2 — Stop building embankments by default. Six decades and 5,000 km have not solved the problem, and in places have deepened it. Where embankments are retained they need real maintenance; where the floodplain can be given room, give it room.
Pillar 3 — Protect the rail and road lifelines. With 38.9% of rail and 17.9% of road in the floodplain, the Northeast’s surface connectivity has a single point of failure. Raising and armouring the most exposed segments is national-security infrastructure, not a state road project.
Pillar 4 — Plan for the four extreme districts. South Salmara Mankachar, Majuli, Barpeta and Dhubri are above 78% exposure. Conventional flood-proofing does not scale to a district where nine buildings in ten are in the water. These need elevated construction standards, raised platforms and community shelters as the default.
Pillar 5 — Basin-scale early warning, including the Barak. Assam floods from two independent systems and from upstream in Arunachal and Tibet. Warning has to be basin-wide, and Silchar’s experience shows the Barak cannot be an afterthought.
The honest bottom line: Assam is the most flood-exposed region in this atlas, and the number is not a modelling artifact — 43.7% of buildings in the modeled 1% floodplain sits right alongside the official 39.58% of land that India has recognised for fifty years. The water is not unusually deep at 1.45 m; it is unusually everywhere. Four districts exceed 78% exposure, Majuli approaches 89% while physically eroding away, and nearly two-fifths of the state’s railway — the Northeast’s land link to the rest of India — stands in the floodplain. Assam’s flood is not an event that arrives. It is the condition the state is built inside.
Sources
- Flood-prone area and official assessment — Government of Assam, Water Resources: Flood & Erosion Problems · Government of Assam: River System of Assam · Drishti IAS: Assam Floods and Flood Management in India
- Embankments and erosion — Sentinel Assam: Ageing embankments and erosion aggravate the flood problem · Scroll.in: Flood-control methods deployed in Assam have actually worsened the problem · PreventionWeb: What’s really behind Assam’s worsening floods?
- 2022 floods and Silchar — Wikipedia: 2022 Silchar Floods · Outlook India: 55 lakh affected as Brahmaputra and Barak inundate new areas · Peoples Dispatch: Flooding in Assam causes massive destruction
- Wider context — Mongabay India: Assam floods expose a crisis beyond heavy rain
- Exposure statistics — India Flood Atlas (modeled 100-year / 1% annual-chance event), NEER.
Exposure figures are modeled estimates (~30 m) for a 1% annual-chance flood, intended for awareness and prioritisation — not parcel-level certainty. Building footprints are structures, not people.