Physical climate risk assessment
Asset-level physical climate risk screening across flood, bushfire, coastal recession and extreme heat, using Australian open hazard data and CMIP6 scenarios.
Connector
Australia's authoritative address file, used as the entry point to the land spine. Coverage, update cadence, licence and the queries the agent runs against it.
G-NAF is the authoritative register of Australian physical addresses. It is the first hop for almost every question Zenancy answers, because a portfolio arrives as text and a hazard layer expects geometry.
Around 15 million address records with coordinates, locality, state, postcode, and links to street and locality tables. Each record carries a reliability code describing how precisely it was geocoded — property centroid, street level, or locality level — which propagates through to the confidence flag on a risk assessment.
| Field | Notes |
|---|---|
| Coverage | All states and territories |
| Records | ~15.3 million addresses |
| Update cadence | Quarterly (February, May, August, November) |
| Licence | Geoscape end user licence, open data, attribution required |
| Custodian | Geoscape Australia |
The quarterly release is loaded into parquet on R2, partitioned by state, with the address detail, street locality and locality tables kept separately and joined at query time. That keeps a state-scoped address lookup to a single partition read.
Geocode this CSV of 240 site addresses and tell me which ones failed to match.
| Result | Count | Reliability |
|---|---|---|
| Matched, property centroid | 228 | High |
| Matched, street level | 9 | Medium |
| Unmatched | 3 | — |
g_nafA G-NAF coordinate is a point. Hazard layers are polygons, and the question "is this site in the flood planning area" depends on the parcel, not the point. So a match here is immediately followed by a cadastral parcel join — the address gets you to the right parcel, and the parcel is what gets intersected with hazard geometry.
Unmatched addresses are reported as unmatched rather than snapped to the nearest guess. You get a list to correct, and the assessment records how many sites were excluded.
Yes. G-NAF is released as open data under the Geoscape end user licence agreement, which permits commercial use with attribution.
Asset-level physical climate risk screening across flood, bushfire, coastal recession and extreme heat, using Australian open hazard data and CMIP6 scenarios.
How the Zenancy agent works — a reasoning loop with a sandboxed Python and DuckDB environment, a parquet data lake of Australian open data, and outputs that always cite their source.
Zenancy is in private preview with Group 2 reporters and their advisers.
Request access