← Insights
Delhi 15 August 2026 · 7 min read · NEER

Delhi and the 1% Flood

A river reclaiming its floodplain, and a city that drowns in its own drains

On 13 July 2023, the Yamuna at Delhi’s Old Railway Bridge climbed to 208.66 m — more than three metres above the 205.33 m danger mark, and 1.17 m above the 1978 record that had stood for forty-five years. Water that a city of twenty million had learned to treat as a distant ribbon came up the storm drains and over the embankments and into places that had not seen a flood in a generation: ITO, Rajghat, the Ring Road, Civil Lines, Yamuna Bazar, the Supreme Court’s approach roads. More than 25,000 people were evacuated; entire east-Delhi colonies and the low-lying belt along the river went under two-and-a-half metres of water that took days to recede. The Yamuna had, briefly, reclaimed its floodplain.

Delhi’s flooding is not a coastal sheet or a hill torrent. It is a two-mechanism, capital-city problem: a great north-Indian river swelling against a city built into its floodplain, and a dense metropolis that waterlogs itself every monsoon because its drains no longer drain. This piece uses the India Flood Atlas — a modeled 1% annual-chance flood layer — to show the scale, explains why nearly half the city sits in harm’s way, and lays out a realistic, phased path to resilience.

The scale

The 2023 flood is observed history. The atlas adds the modeled 1% annual-chance (100-year) flood — the severe event with a 1% chance of arriving in any given year, mapped consistently across the National Capital Territory. Delhi is small enough to model as a single unit, and the result is stark.

Across the NCT, that event reaches:

That 45.4% is the number to sit with. It is not that Delhi floods everywhere at once — the 2023 event did not put half the city underwater. It is that half the city is built low enough, and close enough to the river and its drains, to be reachable by a severe flood. And it floods deep — a third of exposed buildings facing more than two metres — because a river confined between hard urban embankments stacks water high rather than spreading it thin.

A note on what this is: these are modeled figures at ~30 m resolution — regional hazard screening, not a parcel-level guarantee, and building footprints, not people. For a dense city, the ~30 m riverine-and-pluvial model under-captures the fine-grained drainage failure — the blocked nala, the bowl-shaped underpass — that causes much of Delhi’s real flooding. So 45.4% is best read as a floor, not a ceiling.

Where it concentrates

Delhi is one administrative unit, but its flood risk is not evenly spread — it pools in two distinct geographies.

The Yamuna floodplain — the “O-zone.” The river enters at Wazirabad and leaves at Okhla, and between them the Delhi Master Plan designates a roughly 9,700-hectare (~165 km²) ecologically sensitive floodplain — Zone O — meant to stay unbuilt so it can absorb floodwater and recharge groundwater. It has not stayed unbuilt. A DDA survey found that about 7,360 hectares — roughly two-thirds of the O-zone — had been encroached, by colonies, farms, silt dumps and informal settlements from Jharoda and Wazirabad to Sangam Vihar and Milan Vihar. This is the deep-water core: east Delhi and the trans-Yamuna belt, Yamuna Bazar, Majnu ka Tila, Geeta Colony, Sarai Kale Khan. When the river rises, this is what goes under first and deepest — and it is exactly where the floodplain was never supposed to be inhabited.

The drains — the pluvial city. Delhi’s second flood does not come from the river at all. It comes from choked stormwater drains during ordinary monsoon downpours. The Najafgarh drain, the city’s largest natural drainage line, carries the runoff of west and south-west Delhi to the Yamuna — narrowed, silted and encroached along much of its length. And the emblem of the failure is Minto Bridge, a 90-year-old underpass built as a deliberate bowl to preserve a sightline, which floods with a few hours of rain and has repeatedly drowned people and vehicles — a driver in July 2020, and again in the 28 June 2024 deluge that killed eleven across the city. These are not river floods; they are a drainage system that has lost its design capacity.

The 2023 flood is where the two geographies met: a record river backing up into a city whose drains could not push water out against it.

Why Delhi floods

  1. A river confined and encroached. The Yamuna’s floodplain — its natural room to spread — has been built over across two-thirds of the O-zone. A river denied its floodplain rises higher for the same volume, and it rises into the very settlements that took the floodplain’s place.
  2. Upstream releases arrive fast. Delhi’s flood peaks are largely made in Haryana. When the Hathnikund Barrage releases — around 359,000 cusecs in the days before 13 July 2023 — the pulse reaches Delhi in a day or two, with little the city can do but evacuate the low belt.
  3. Silted, encroached drains. The Najafgarh drain and the city’s stormwater network no longer carry their design flow. Runoff that should leave the city instead ponds in it.
  4. A built-over, impervious catchment. Concrete and construction have replaced the soil and water bodies that once absorbed rain, so more runoff arrives faster at drains that can carry less.
  5. Bowl-shaped chokepoints. Underpasses like Minto Bridge and dozens of others sit below the water table of a heavy shower, turning roads into traps within minutes.
  6. Backwater lock-in. When the Yamuna is high, the city’s drains cannot discharge into it — so river flood and drain flood compound each other, exactly the 2023 pattern.
  7. Climate change is intensifying the cloudburst-scale rainfall that drives both the upstream river pulses and the local urban deluges.

A strategic approach

Delhi’s problem splits cleanly, and so should its strategy: manage the river and protect its floodplain, and rebuild the capacity of the drains.

Pillar 1 — Know the risk, and warn early. Delhi already has a functioning flood-warning chain tied to Hathnikund releases — the day or two of travel time is real lead time. Sharpen it with the atlas’s exposure map so evacuation targets the deepest O-zone belt first, and extend nowcasting for the pluvial floods that give no upstream warning at all.

Pillar 2 — Reclaim and protect the floodplain. The single highest-leverage act is to stop and reverse encroachment of the O-zone. Every hectare of floodplain restored is storage the next flood can use, and every settlement kept out of it is people not evacuated. This is a governance-and-enforcement fight more than an engineering one.

Pillar 3 — Unclog and remodel the drains. Desilt, widen and de-encroach the Najafgarh drain and the stormwater network before every monsoon, and fix the bowl-shaped underpasses with pumps, sumps and monitored closures. This is the lowest-glamour, highest-return work Delhi can do, and it saves lives in the next monsoon.

Pillar 4 — Coordinate upstream. Delhi’s peaks are set at Hathnikund and moderated (or not) by barrage operation across Haryana, Uttar Pradesh and Delhi. Inter-state, basin-scale coordination of releases and real-time forecasting is what converts a wall of water into a managed one.

Pillar 5 — Govern land use. Enforce the O-zone as a no-build floodplain, protect the remaining water bodies and drainage lines, and require flood-safe design in the exposed belt. The cheapest flood control is not building in the floodway.

Pillar 6 — Make buildings and communities resilient. For the roughly half-million exposed footprints already in place — especially in the deep east-Delhi belt — raised plinths, protected utilities, rehearsed evacuation, and flood-inclusive home insurance (covered under “STFI” perils in standard Indian cover such as Bharat Griha Raksha, yet rarely held).

A realistic timeline

Flood resilience is a 15–25 year programme, not a five-year project — and undoing decades of floodplain encroachment is measured in political will as much as budget.

Phase 0 — Foundation & quick wins (0–2 years). Low cost, high impact. Exposure-targeted evacuation planning for the O-zone belt; desilt and de-encroach the Najafgarh drain and city stormwater network before each monsoon; pump-and-close protocols for the killer underpasses; sharpen the Hathnikund warning chain.

Phase 1 — Reclaim & unclog (2–5 years). Halt and roll back O-zone encroachment; remodel the drainage network to its design capacity; enact and enforce floodplain zoning; stand up a permanent annual maintenance budget so cleared drains stay clear.

Phase 2 — Basin-scale resilience (5–10 years). Move from drain-by-drain fixes to whole-catchment drainage, and from city-only to inter-state Yamuna basin management — coordinated barrage operation and real-time forecasting. Retrofit exposed public assets (hospitals, schools, substations, metro, rail) in the flood belt.

Phase 3 — Climate-proof & sustain (10–20+ years). Size drainage, embankments and barrage rules for 2050/2100 rainfall (the SSP scenarios), and lock in maintenance, monitoring and updating as permanent functions.

The honest bottom line: Delhi exposes a larger share of its buildings — 45.4%, nearly half — than any other state in the atlas, and it floods deep, because a national capital has been built into its river’s floodplain and drained through channels it has let choke. And even that number is a floor: a ~30 m model under-captures the drain-by-drain waterlogging that drowns people at underpasses in an ordinary shower. Two problems, two fixes — give the river its floodplain back, and give the drains their capacity back. Neither is exotic engineering. Both are governance. Knowing exactly where the deepest water sits is what lets a limited budget protect the most people.

Sources

Exposure figures are modeled estimates (~30 m) for a 1% annual-chance flood, intended for awareness and prioritisation — not parcel-level certainty.

← All insights