Britain's flood exposure atlas

Flood hazard, exposure, vulnerability and risk: four different quantities

Hazard describes the probability, extent, intensity or depth of a flood event. Exposure is the presence of people, buildings, land parcels and assets within that hazard footprint. Vulnerability is the susceptibility of those assets to damage. Risk combines hazard, exposure, vulnerability and relevant consequence pathways. A coloured zone on a map is not, on its own, a validated estimate of annual property damage or insurance cost.

Annual exceedance probability (AEP)

A 1% AEP flood corresponds to a 1-in-100 annual chance under the modelled conditions; it is not a prediction of a flood arriving once every 100 years. The chance of at least one event in a 30-year period under the simplifying assumptions of stationary, independent annual probabilities is 1 − (1 − 0.01)³⁰ ≈ 26%. Climate change and non-stationarity undermine simple extrapolation over long asset lifetimes.

Flood Zone 2, Flood Zone 3 and defended risk

Planning flood zones and risk-to-properties products serve different statutory and analytical purposes. The Environment Agency's Flood Zones plus Climate Change layer models the combined extent of zones 2 and 3 under future allowances without credit for existing flood-defence benefits; RoFRS considers the presence and condition of defences. These must not be labelled equivalent probabilities. [1,4,5]

Surface-water flood risk and drainage capacity

Surface-water maps identify land susceptible to intense rainfall before water enters watercourses. England's planning extent datasets cover 3.3%, 1% and 0.1% AEP scenarios, with complementary depth bands. They are area-level decision-support products, not proof of entry into an individual building. [2,3]

Official flood exposure GIS datasets and variable definitions

The register distinguishes original publishers, intended use and modelling limitations. It is a downloadable source dictionary, not an invented set of modelled exposure counts.

Geographic coverage and data comparability

England: Environment Agency datasets are the foundation of the detailed comparison. Wales: Natural Resources Wales publishes its own flood maps and planning datasets. Scotland: SEPA flood maps apply distinct national methodology. Northern Ireland: Department for Infrastructure flood mapping follows its own conventions. Each nation requires its own source extraction, licence review and methodological crosswalk before any combined British or UK ranking. [6–9]

A reproducible workflow for calculating flood-exposed sites

1. Specify estimand. Decide whether the target is land area (hectares), residential addresses, dwellings, population, roads or development sites. Do not treat one as a proxy for another without validation.
2. Fix the geography and observation date. Obtain versioned flood polygons, official boundary datasets and appropriately licensed building or address geographies for one reference date.
3. Separate hazards. Create distinct fluvial, coastal and pluvial indicators at matched AEP and climate horizon; preserve defence assumption as an explicit field.
4. Perform spatial intersection. Reproject layers consistently, validate geometries, detect polygon overlap and intersect with parcels or building footprints. Keep a unique feature identifier and a record of missing geometries.
5. Calculate exposure measures. Report number and percentage of intersecting assets as well as flood-zone area, with clear selection rules for partial intersection, building centroid and uncertainty.
6. Test sensitivity. Compare baseline with climate scenarios; vary AEP, parcel-intersection threshold, spatial resolution and defence treatment. Avoid double-counting a site exposed to multiple hazards.
7. Audit and publish. Supply the extract date, version, boundary year, GIS software, code, licence restrictions and source URLs alongside numerical tables.

Suggested analysis schema

Why flood layers cannot simply be added together

Flood mapping is not a single uniform dataset. The relevant question might concern the legal-planning extent of a river floodplain, the annual likelihood of water reaching a location under contemporary conditions, the depth associated with extreme rainfall, or a future scenario incorporating a changing climate. Those questions have different modelling inputs, outputs and intended audiences. A robust atlas must keep them as separate measurements before attempting any form of multi-hazard synthesis.

Four dimensions of a defensible comparison

A national visualisation can show these products next to each other, but a composite index needs explicit weighting, common geography and a declared decision purpose. Weighting hazard maps without a defensible conceptual model risks making the index less transparent than its constituent layers.

The problem of defence assumptions

Many planning datasets describe the floodplain that would be affected under a stated hazard scenario irrespective of some existing protection, while other datasets estimate the likelihood of flooding with defence systems represented. These are both useful outputs. The former can help determine whether development should be considered in a location; the latter can support a different assessment of actual defended exposure. It would be methodologically incorrect to label them two versions of the same metric and subtract one polygon area from another to estimate a precise number of protected homes without aligned models and a specified asset layer.

Climate projections and development lifetime

Environment Agency allowances address changes to peak river flows, peak rainfall intensity, sea level and coastal forcing over different future periods. They are based on climate projections and differ by geographical context and development characteristics. An assessment should specify which allowance was selected and why, rather than applying an arbitrary national percentage uplift to every flood surface. The relevant future horizon also depends on the anticipated lifetime and vulnerability of a proposed development. [10]

Climate uncertainty is not resolved by selecting a single visually alarming layer. A careful analysis should present a central planning case and relevant higher sensitivity cases, documenting the differences between them. A mapped future flood extent is neither a realised event in 2026 nor a statement that every included property will flood during its lifetime.

How flood exposure data informs property due diligence and planning

Planning consultants can use statutory flood maps to identify whether an application may require a site-specific Flood Risk Assessment (FRA) and sequential testing, subject to the relevant national policy and local planning context. Surveyors and buyers can use these layers to scope professional investigations; insurers require more granular information than a flood-zone overlay. The Environment Agency explicitly warns that mapped products cannot determine flood risk to a particular property. [1–5]

Climate change and development lifetime

Flood Zones plus Climate Change includes future fluvial and coastal scenarios. The mapped extent depends on modelling assumptions, sea-level rise, chosen allowance and planning horizon. It is a scenario analysis rather than an observed 2026 flood footprint or prediction that every mapped site will flood. [1]

Five key research limitations

National comparability is constrained by differing models across UK nations; map cells are not property-level surveys; flood defences may fail or be overtopped; spatial resolution and incomplete local drainage information affect uncertainty; and new development, changing rainfall, land use and exposure mean a snapshot cannot establish future annual loss. This atlas does not publish unsupported borough-level flood rankings or property counts.

Questions for subsequent empirical research

What share of homes and consented developments intersects 1% AEP surface-water footprints? How much does inclusion of future flood extents alter local planning-constraint statistics? Do lower-income neighbourhoods have a greater proportion of exposed building footprints once settlement density is controlled? Such questions require licensed, harmonised asset data and reproducible statistical analysis before quantitative conclusions can be drawn.

From a flood map to a defensible exposure estimate

Several tractable questions follow from the dataset register, but answering them rigorously requires asset data, consistent spatial definitions and a transparent treatment of missing records. This publication sets out the research design and does not present speculative headline estimates as if they had been calculated.

Candidate outcome measures

1. Building-footprint intersection. The share of licensed residential footprints intersecting selected baseline flood extents, reported separately for river/coastal and surface-water hazards. Specify any overlap threshold and examine centroid-based sensitivity.
2. Future-exposure increment. The additional buildings or land area within selected future scenario polygons versus a specified baseline. Count unique assets and report the precise climate scenario, allowance and comparison period.
3. Planning-site exposure. The proportion of identified housing allocation sites and submitted planning sites crossing hazard polygons. Record development status, boundary quality and whether a formal flood-risk assessment exists.
4. Socio-spatial distribution. Compare exposed residential assets by deprivation or other local context while controlling for settlement structure and avoiding ecological conclusions about individual residents.
5. Risk-management sensitivity. Contrast results under defended and undefended assumptions, with clear uncertainty statements instead of claiming that a map alone establishes protection or expected annual financial loss.

Minimum reporting standard for a replicated atlas

Each published results table should name the source agency, extraction timestamp, licensing conditions, boundary reference date, spatial reference system, hazard and scenario identifiers, asset denominator, overlay rule, area of missing coverage, deduplication strategy and data revision history. Releasing a machine-readable variable dictionary alongside the tables would make comparisons auditable and allow readers to identify genuine changes in exposure separately from changes in source coverage.

Where uncertainty is high, report bounds or sensitivity cases. A zero in an unmeasured local authority should never be substituted for missing information. Small differences between authority percentages should not be described as meaningful without evaluating the resolution and completeness of the contributing layers.

Implications for a site-specific decision

For an individual site, national spatial layers should begin an investigation rather than end it. A suitable professional appraisal may require local strategic flood-risk assessment material, site levels, the history and condition of defences, drainage records, relevant catchment information, recent flood events and consultation with the local planning authority or relevant flood-risk body. The appropriate evidence varies with the proposed land use and local jurisdiction. [6–9,11–12]

For research communicators, the practical test is whether a reader can identify what was actually observed or modelled, what Maplify calculated, and what remains unknown. That separation is more useful than presenting a single unqualified national risk score.

Common analytical errors when interpreting a flood-risk map

Flood products are often shown as continuous coloured areas, creating a misleading appearance of uniform certainty. Yet the input models, terrain resolution, probability assumptions and condition of defences vary. Using a single map screenshot as proof of a property's exposure removes the dataset from the definitions necessary to interpret it. Four recurring errors deserve explicit scrutiny.

Conflating probability with recurrence interval

A 1% annual exceedance probability does not mean one flood will occur exactly once in every century, or that a property is “safe for 99 years”. Even in a hypothetical stationary independent annual model, the probability of at least one event increases over a longer observation period. The calculation is 1 − (1 − p)ⁿ, where p is the stated annual probability and n the number of years. This is a mathematical illustration of event probability, not a model of actual losses to a building.

These percentages are derived from the stated toy probability model. They must not be interpreted as a future flood forecast for a particular address, and an assumed constant annual probability will generally not describe a changing climate, drainage catchment or flood-defence system.

Counting one site twice under multiple hazards

Where flood-risk polygons overlap, adding site counts across sources can inflate the number of distinct exposed sites. For example, a riverside parcel may intersect both a river inundation model and a surface-water hotspot. An evidence-based multi-hazard study should report each hazard separately and a deduplicated “any hazard” union; the simple sum is a count of intersections rather than unique parcels. This requires common boundary identifiers and consistently handled multipart polygons.

Comparing different UK administrations without a crosswalk

England’s Environment Agency maps, Natural Resources Wales data, Scotland’s SEPA products and Northern Ireland’s Department for Infrastructure maps are not one harmonised national modelling release. Individual agencies publish different probability scenarios, mapping specifications and access arrangements. The presence of a polygon from one nation does not mean that an identically coloured polygon in another nation represents the same analytical variable. [6–9]

Confusing mapped hazard with expected building damage

Damages depend on whether water reaches a building, depth, duration, velocity, construction details, contents and existing protection. A mapped hazard at the parcel scale is insufficient to estimate household insurance premiums or expected annual financial damage. Researchers would need defensible building footprints, threshold heights, damage functions and hazard-frequency curves before estimating losses, with the uncertainties of each layer propagated through the result.

A practical route to verified local data

To move from these national datasets to a specific site, use the Maplify free flood-risk map for initial geographic context, then verify the official flood zones and local evidence from the relevant national agency and planning authority. For a formal planning application, assess the current local flood-risk policy and determine whether a professional Flood Risk Assessment is needed. The map is a screening aid and does not replace official consultation or a measured ground-level assessment. [6,10,11]

References, original datasets and suggested citation

1. Environment Agency — Flood Zones plus Climate Change (England, updated October 2026)www.data.gov.uk
2. Environment Agency — Surface Water Spatial Planning Extents: Climate Changewww.data.gov.uk
3. Environment Agency — Surface Water Spatial Planning Depths: Climate Changewww.data.gov.uk
4. Environment Agency — Risk of Flooding from Rivers and Sea: Climate Changewww.data.gov.uk
5. Environment Agency — defended and undefended flood risk extentswww.data.gov.uk
6. GOV.UK — long-term flood risk information (England)www.gov.uk
7. Natural Resources Wales — flood maps and flood risk informationnaturalresources.wales
8. Scottish Environment Protection Agency — flood mapswww.sepa.org.uk
9. Department for Infrastructure Northern Ireland — flood mapswww.infrastructure-ni.gov.uk
10. Environment Agency — flood risk assessments: climate change allowanceswww.gov.uk
11. GOV.UK — flood risk assessments for planning permission (England)www.gov.uk
12. NPPF — managing flood risk and coastal change (England)www.gov.uk

Suggested citation: Maplify Research Desk (2026). Britain's flood exposure atlas: a methodological review of national flood hazard datasets and planning applications. Maplify Research, 27 October 2026. Available at: https://maplify.co.uk/research/flood-exposure-atlas (accessed on date of use).

Source and usage notice: Links point to official publisher records. Data licensing varies by product and nation. Verify the current layer date and permitted reuse on each publisher site. The study is educational and does not replace an Environment Agency, SEPA, NRW or DfI site-specific assessment, professional survey, or statutory planning consultation.