Methodology
How radon.tips calculates radon risk by postcode district.
Source data
BGS Radon Indicative Atlas v3
The British Geological Survey (BGS) publishes the Radon Indicative Atlas as a GeoPackage (.gpkg) file. It contains approximately 320,000 1km² grid squares across Great Britain, each assigned a CLASS_MAX value from 1 to 6:
| Class | Label | % Homes above Action Level |
|---|---|---|
| 1 | Low | <1% |
| 2 | Moderate | 1–3% |
| 3 | High | 3–5% |
| 4 | Very High | 5–10% |
| 5 | Highest | 10–30% |
| 6 | Highest | >30% |
The "Action Level" is 200 Bq/m³ — the UK Health Security Agency threshold above which action is recommended.
ONS NSPL February 2026
The ONS National Statistics Postcode Lookup (NSPL) contains grid references (OSGB36 easting/northing) for approximately 2.6 million UK postcodes. We use February 2026 edition, filtered to England (country code E92000001).
Processing pipeline
Step 1: Grid tile lookup
Each BGS grid square occupies a 1km × 1km cell in the OSGB36 national grid. We extract the SW corner of each cell (rounded to the nearest 1,000m) to create a tile lookup dictionary: (easting_floor, northing_floor) → CLASS_MAX.
Step 2: Postcode matching
For each postcode in the NSPL, we floor its easting and northing to the nearest 1,000m and look up the matching BGS tile. A small proportion of postcodes have no BGS match (typically coastal or non-residential areas) and are excluded.
Step 3: District aggregation
Postcodes are grouped by their outward code (district), e.g. all EX1 postcodes form one district. For each district, we compute:
- Modal class (risk_class) — the most commonly occurring BGS class among postcodes in the district. This is used as the district's representative risk level.
- Maximum class (max_class) — the highest BGS class found in any postcode in the district.
- % high-risk (pct_high_risk) — the proportion of postcodes with class ≥ 3.
- Centroid — mean easting/northing converted to WGS84 latitude/longitude.
- Grid square count — the number of distinct 1km tiles present in the district.
Step 4: County aggregation
Districts are grouped by county using the NSPL's county code field (cty25cd), joined to ONS county names from the Dec 2020 County Names and Codes dataset. The county's representative risk is the mean of its districts' modal classes, rounded to the nearest integer.
Note: Unitary authorities (including London, Greater Manchester, and many others) have no county code in the NSPL and are grouped as "Unknown". These areas are excluded from county-level pages but all their postcode districts appear individually.
Risk class labels
The six BGS classes are mapped to display labels as follows. Classes 5 and 6 share the "Highest" label, as the distinction is primarily geological:
| Class | Label | % Homes above Action Level |
|---|---|---|
| 1 | Low | <1% |
| 2 | Moderate | 1–3% |
| 3 | High | 3–5% |
| 4 | Very High | 5–10% |
| 5 | Highest | 10–30% |
| 6 | Highest | >30% |
Limitations and caveats
- The BGS Indicative Atlas is a statistical model, not a direct measurement. Actual radon levels in individual homes vary due to building type, ventilation, occupant behaviour, and local geology.
- The modal class gives the most representative figure for a postcode district, but individual postcodes within a district may differ significantly.
- Postcode districts vary enormously in size and postcode count. Small or rural districts may have fewer measurements contributing to the modal class.
- Terminated postcodes (postcodes no longer in active use) are excluded from the analysis.
- The dataset is based on the February 2026 NSPL edition. Districts created or substantially altered since this date may not be represented accurately.
- County names are sourced from the ONS Dec 2020 County Names and Codes dataset. County boundaries are assumed stable between Dec 2020 and the Feb 2026 NSPL edition; any boundary changes in this period may cause a small number of districts to be assigned to incorrect counties.
Open data licences
BGS data © British Geological Survey (UKRI). Contains public sector information licensed under the Open Government Licence v3.0.
ONS NSPL © Crown copyright and database right 2026. Contains public sector information licensed under the Open Government Licence v3.0.