What is flood risk data?

Flood risk data is the set of maps and grids that describe where flooding is likely, how deep the water would be, and how often it is expected. It comes as regulatory flood zones, such as FEMA’s Special Flood Hazard Area, and as modeled depth grids for return periods like the 100-year and 500-year flood.

What flood risk data is

Flood risk data expresses flood probability in space. The defining property is a return period: a 100-year flood zone is the area with a 1 percent chance of flooding in any given year, and a 500-year zone has a 0.2 percent chance.

In the United States the regulatory source is FEMA’s National Flood Hazard Layer (NFHL). It maps Zones A and AE (1 percent riverine and coastal), Zone V (coastal with wave action), and Zone X (0.2 percent or minimal risk) as vector polygons.

Modeled flood risk data goes further. Providers such as First Street, Fathom, and JBA produce depth grids at 3 to 30 m resolution for several return periods, covering pluvial, fluvial, and coastal flooding, including areas FEMA has never mapped.

Global public sources include the European Commission’s JRC Global Flood Hazard Maps, with depth grids for return periods from 10 to 500 years at about 1 km.

Observed flood extent is a third class, mapped from satellite imagery after an event.

What flood risk data does

Flood risk data converts a location into a probability and a depth, which is the input every flood decision needs.

Lenders require flood insurance on any mortgaged property inside a FEMA Special Flood Hazard Area, so the zone sets a cost for millions of homeowners.

Insurers price policies and measure portfolio exposure by intersecting insured locations with zones and depth grids, then modeling loss from depth-damage curves.

Real estate investors screen acquisitions and rank a portfolio by expected flood loss under current and future climate scenarios.

Cities and utilities plan drainage, elevate substations, and site critical facilities from depth grids.

Emergency managers pre-position crews by overlaying forecast rainfall on pluvial flood risk data.

Each of these decisions is a spatial join between flood risk data and the assets it concerns.

How flood risk data works

Flood risk data is produced by hydrologic and hydraulic modeling. A digital elevation model supplies the terrain, rainfall or river gauge statistics supply the water, and a hydraulic model routes that water across the terrain to estimate depth and extent for each return period.

Data type	Form	Example	Resolution
Regulatory zones	Vector polygons	FEMA NFHL	Parcel scale
Modeled depth grids	Raster, one per return period	First Street, Fathom, JRC	3 m to 1 km
Observed extent	Vector or raster from imagery	Copernicus EMS, Sentinel-1	10 to 30 m
Historical claims	Points or tabular	NFIP redacted claims	Address

Zones answer a binary question: in or out. Depth grids answer a continuous one: how deep, and for which return period, which is what a loss model needs.

Resolution sets usefulness. A 1 km global grid places whole neighborhoods in or out; a 3 m grid distinguishes a house on a rise from its neighbor in the swale.

Currency matters as much. FEMA maps are decades old in many counties, and climate-adjusted datasets model 2050 rainfall and sea level.

FEMA delivers through downloads and ArcGIS Feature Services; commercial providers deliver licensed GeoTIFF or GeoParquet packages.

The output of a flood risk analysis is a table of assets with a zone code, a depth per return period, and an expected loss.

Flood risk data in Wherobots

WherobotsDB joins flood risk data to parcels, buildings, and policies in one SQL query, whether the flood layer is a vector zone or a raster depth grid.

The ArcGIS Feature Service reader pulls FEMA’s National Flood Hazard Layer directly from the FEMA REST endpoint, with filter and column pushdown. Zone polygons join to property points or parcel polygons with ST_Intersects.

Depth grids load as out-db rasters with RS_FromPath. RS_Value returns the depth at each property point, and RS_ZonalStats returns the maximum depth inside each parcel or footprint, per return period, in the same query.

The California coastal flood risk notebook in Wherobots Cloud runs this pattern end to end. It intersects parcels with FEMA zones and sums assessed value at risk per city.

A spatial join across millions of policy locations and every zone polygon in a state completes on a single runtime, because WherobotsDB partitions both sides spatially.

Observed flood extent from Sentinel-1 SAR comes from the same engine, so a post-event count of flooded buildings compares directly against the pre-event modeled exposure.

Data stays in the customer’s cloud storage or Iceberg catalog. The Wherobots MCP Server exposes the same tables and functions to Claude Code and other agents, so an underwriter can ask for exposed policies in a county in plain language.

-- Policies inside a FEMA Special Flood Hazard Area plus modeled 100-year depth
SELECT pol.policy_id, z.FLD_ZONE,
       RS_Value(d.rast, pol.geometry) AS depth_100yr_m
FROM policies pol
JOIN fema_nfhl_zones z                  -- read via the ArcGIS Feature Service reader
  ON ST_Intersects(z.geometry, pol.geometry)
LEFT JOIN flood_depth_100yr d           -- out-db GeoTIFF depth grid
  ON RS_Intersects(d.rast, pol.geometry)
WHERE z.FLD_ZONE IN ('A', 'AE', 'AO', 'AH', 'V', 'VE');

Related terms

• Parcel data: the property layer flood zones are intersected with
• Digital elevation model: the terrain input every flood model starts from
• SAR data: how observed flood extent is mapped after an event
• Catastrophe modeling: where flood risk data feeds insurance loss estimates

Read more from Wherobots

• California coastal flood risk: property exposure analysis
• Insurance & Risk use case
• El Niño 2026, atmospheric rivers, and California’s burn scars: one query, two data models
• How to score every building in a state for catastrophe risk

Intersect your portfolio with flood risk data in WherobotsDB on the free tier at cloud.wherobots.com.