Priority Habitat Inventory (PHI)
What it is
The Priority Habitat Inventory (PHI) is Natural England's spatial map of England's priority habitats — the habitats listed as being of principal importance for conserving biodiversity under Section 41 of the Natural Environment and Rural Communities (NERC) Act 2006, and originally identified through the UK Biodiversity Action Plan.
It currently maps around 27 terrestrial and freshwater priority habitat types — deciduous woodland, lowland meadows, lowland calcareous grassland, coastal and floodplain grazing marsh, blanket bog, and so on — as polygons across the whole of England. In practice it is the closest thing England has to a single authoritative "where are the important habitats" layer, and it is one of the highest-priority inputs to a BNG baseline.
How it's produced
This is the part that matters most, and it is where PHI's character differs sharply from machine-classified land cover.
PHI is compiled and interpreted, not remotely classified. Habitat boundaries are largely captured against Ordnance Survey MasterMap (OSMM) polygons, then populated from a wide range of underlying evidence: field survey, Sites of Special Scientific Interest data held in Natural England's ENSIS database, NVC (National Vegetation Classification) survey codes, and other local and national survey sources. Polygons are merged or split to form coherent habitat patches.
The consequence: where a PHI polygon exists and is well-sourced, it typically reflects human ecological judgement grounded in survey — a genuinely different (and generally higher) confidence profile than a probability output from a classifier. But that strength is uneven: PHI is a compilation of many sources of varying age and rigour, so confidence is not uniform across the map, and the layer does not attach a per-polygon numerical confidence score the way an ML-derived product does.
Update frequency & currency
Natural England publishes updates to PHI on a roughly twice-yearly cycle. That cadence is better than most public habitat datasets, but "updated twice a year" does not mean "every polygon is current" — an update pass incorporates newly available survey and corrections; it does not re-survey England. Individual polygons can rest on evidence that is many years old. Treat the publication date as the date the compilation was refreshed, not the date the ground was last checked.
Spatial resolution / precision
Because boundaries are captured against OSMM, PHI inherits OS MasterMap's fine polygon geometry — spatially precise lines on the ground. The precision to be careful about is thematic, not geometric: a crisply drawn boundary can still enclose a habitat classification that is coarse, generalised, or out of date. Sharp edges are not the same as sharp evidence.
Known limitations
- Coverage gaps. PHI maps priority habitats. A parcel with no PHI polygon is not necessarily habitat-free — it may simply be habitat that PHI never captured. Absence of a polygon is not evidence of absence of habitat.
- Uneven, undated confidence. Sources and vintages vary polygon to polygon, with no per-feature reliability score to tell them apart.
- England only. No Welsh, Scottish, or Northern Irish coverage — those nations have their own habitat datasets.
- Thematic staleness. Land use changes faster than the inventory. A polygon mapped as a priority grassland years ago may since have been ploughed or built on.
How it compares to Living England
PHI and Living England are the pairing most often confused, because they can produce superficially similar polygons over the same ground while meaning very different things.
| Priority Habitat Inventory (PHI) | Living England | |
|---|---|---|
| Method | Compiled/interpreted from survey, ENSIS, NVC, against OSMM | Machine-learning classification of satellite imagery |
| Evidence type | Human ecological judgement, ground-truthed in part | Modelled probability from remote sensing |
| Per-feature confidence | No numerical score; varies by source | Carries a machine-learning reliability field (relblty) |
| What it maps | ~27 priority habitat types (a curated subset) | Broad habitat classes across the whole landscape |
| Best treated as | "Where do we have surveyed evidence of valuable habitat?" | "What does the imagery suggest is here, everywhere?" |
The headline: they are different kinds of evidence, not two versions of the same map. PHI answers "where is there surveyed priority habitat?"; Living England answers "what is the most likely land cover class here, based on imagery?". Confusing the two — trusting an ML class as if it were a survey, or dismissing a PHI gap as if it meant "nothing here" — is a classic desk-assessment error.
The methodological contrast above (surveyed/compiled vs. ML-classified, and Living
England's relblty field) is well established and carried from WildStack's
internal dataset documentation. The specific count of "~27 priority habitat types"
should be reconfirmed against the current PHI technical documentation at review
time, as the mapped set has changed across PHI versions.
Role in BNG assessment
PHI is high-priority, near-authoritative input to a BNG baseline. In WildStack's dataset hierarchy it sits at the top for habitat identification and condition: where a PHI polygon exists, its field-derived classification is preferred over lower-confidence proxies, and it directly informs habitat distinctiveness — which is what makes the biodiversity gain hierarchy bite. It is not itself the statutory metric, but it is one of the most trustworthy things you can feed into it.
Official source
- Priority Habitats Inventory (England) — Natural England / data.gov.uk
- Habitats and species of principal importance in England (S41 NERC Act 2006) — GOV.UK
Last reviewed
5 July 2026. This page will need revision if: PHI changes its production methodology or the set of mapped priority habitats, Natural England alters the update cadence, or a per-feature confidence attribute is introduced. Reconfirm the mapped-habitat count and the licence terms against the data.gov.uk record at next review.