Who it's for
Free, citable flood exposure for humanitarian and development organisations — to target programmes, pre-position relief, and evidence funding proposals.
Free, with no login: the full 100-year map for all 36 states and union territories, the 743-district ranking, road and rail exposure, and 36 state analyses — all citable in a proposal. What is not free is other return periods, future-climate scenarios, and bulk or programmatic access. If you are working on flood resilience and the paid tier is out of reach, tell us — we would rather the data got used.
The problem
Targeting decisions — which blocks to work in, where to pre-position stock, which communities to enrol in early warning — are often made from last year’s flood, which is a poor predictor of the next one.
The commercial datasets that would answer this properly are priced for insurers. For most programme budgets they are simply out of reach.
And a proposal that says "this district floods" is weaker than one that says how much of it does, on evidence a donor can independently check.
In practice
Each one is a real workflow — who does it, the situation, the steps, and what comes back.
You have funding for a flood-resilience programme in a state and need to choose districts on something better than where the last emergency was.
What you get: A defensible shortlist of districts, with the reasoning you can put in front of a board.
GET /v1/exposure/admin/state/{code} Warehousing is finite. Stock placed in the wrong district is stock that arrives late — or is itself flooded.
What you get: Placement decisions that survive the event they were made for.
GET /v1/point Under Section 135 a corporate funder must justify where CSR money goes. Disaster resilience qualifies — but the case has to be evidenced.
What you get: An independent, free-to-cite evidence base in a document a compliance team will accept.
People know their village floods. What they usually do not have is a picture of how their area compares, or how deep the water gets.
What you get: Material that can be used in a community meeting without needing a GIS to interpret it.
Early-warning systems work when the right households are enrolled. Coverage is usually decided by administrative convenience.
What you get: Enrolment concentrated where the modeled exposure actually is.
POST /v1/exposure How you get it
GET /v1/exposure/admin/{level}/{code} POST /v1/exposure GET /v1/point Organization and Enterprise are commercial plans — see the API page or talk to us. Return periods other than the 100-year event, and climate scenarios, are not part of the free layer.
Related reading
43.7% of buildings exposed — the largest humanitarian caseload in the atlas
Read the analysis →Why a low flood number does not mean low risk — the landslide blind spot, stated plainly
Read the analysis →Where rain falls is not where water stops — the transboundary case
Read the analysis →This models riverine and pluvial flooding for a 1% annual-chance event. It does not model cyclone storm surge, dam releases, or landslides — and in the hill states landslides, not inundation, are the leading cause of death. Use it to prioritise, and pair it with local knowledge.
We publish the limits next to the numbers. A risk map that hides its own weaknesses isn’t a public good.
If you are working on flood resilience in India and the data would help, we would like to hear from you.
Also for
Flood exposure for planning, not just for response
Read more →Find out before you sign
Read more →Price and accumulate Indian flood risk on Indian terrain
Read more →See the flood exposure sitting inside your loan book
Read more →Know what flood risk is doing to portfolio value
Read more →Keep the plant running — and know what to disclose
Read more →Check the site before the design is fixed
Read more →