Skip to content

LINZ data via eolas

Land Information New Zealand (LINZ) is the agency responsible for cadastral, geodetic, hydrographic, and topographic data. eolas serves 106 datasets from LINZ — covering property parcels, addresses, road centrelines, suburbs, place names, and the geodetic mark networks (including Antarctic + offshore islands).

If you're doing property-data, cadastral, or NZ-wide spatial-reference work, LINZ is the primary source.

What's in the catalogue

Cadastral

Dataset Description
nz_parcels All NZ land parcels (~3 million features). The cadastral foundation layer.
nz_primary_parcels Primary parcels (one per title); ~2 million. Faster than nz_parcels for most use cases.
nz_non_primary_parcels Easements, roads, secondary parcels.
nz_property_titles Title records keyed to parcels.
nz_landonline_parcels Landonline working copy — useful for in-flight subdivisions before they propagate to nz_parcels.

Addresses + place names

Dataset Description
nz_addresses Street addresses (~3 million). LINZ's authoritative address layer.
nz_street_address Address points only (no extra metadata).
nz_suburb_locality Suburb / locality polygons.
nz_suburbs_and_localities The combined LINZ table — what most users want for "suburb of X".
nz_suburb_suburb_locality_territorial_authority Suburb-to-TA crosswalk.
nz_place_names NZGB authoritative place names — official and historical, with iwi/hapū attribution.

Buildings + topography

Dataset Description
nz_building_outlines Building footprints (~2 million).
nz_road_centrelines Road network as line geometry.
nz_river_centrelines River network.
nz_coastlines NZ coastline polygons.

Geodetic + survey

Dataset Description
nz_trig_points Trigonometric survey marks.
antarctic_geodetic_marks Antarctic survey network — NZ's responsibility under the Antarctic Treaty.
antarctic_geodetic_vertical_marks Vertical control marks (heights).
canterbury_eq_geodetic_marks_2010_11, _2016 Marks affected by the 2010-2011 and 2016 Canterbury earthquakes (movement records).

For the full list, browse eolas.fyi/datasets?source=LINZ.

Refresh schedule

Weekly, Wednesday morning NZ time. The cadastral layers (parcels, titles) update frequently as Landonline transactions complete; the topographic and geodetic layers are more stable.

meta = client.info("nz_parcels")
meta["last_refreshed_at"]
meta["source_last_modified_at"]   # LINZ's Koordinates publish timestamp

License

All LINZ data is published under CC-BY 4.0. You can use it commercially, derive from it, and redistribute — with attribution.

Recommended attribution: "Source: LINZ, served via eolas (eolas.fyi). CC-BY 4.0."

LINZ also has specific guidance on derivative works (e.g. for property analytics products) at www.linz.govt.nz/licensing. Worth a look if you're building a commercial product on LINZ data.

Common patterns

Plot all NZ road centrelines

# ~2 million features — needs Pro tier to avoid the 50k row cap, or filter first
roads = client.linz("nz_road_centrelines", as_sf=True)
roads.plot(figsize=(10, 12), linewidth=0.1, color="grey")

library(sf)
library(ggplot2)

roads <- eolas_get_linz("nz_road_centrelines", as_sf = TRUE)
ggplot(roads) + geom_sf(linewidth = 0.05) + theme_void()

Look up addresses near a point

LINZ doesn't ship spatial-index filters in the eolas API yet — you'd download a region first, then filter in memory:

import geopandas as gpd
from shapely.geometry import Point

# Fetch addresses in a bounding box (currently no API-side filter; pull region + filter local)
addresses = client.linz("nz_addresses", as_sf=True)

# Filter to within 500m of a point
pt = gpd.GeoSeries([Point(174.7762, -41.2865)], crs="EPSG:4326")  # Wellington
pt_nztm = pt.to_crs("EPSG:2193")
addr_nztm = addresses.to_crs("EPSG:2193")

nearby = addr_nztm[addr_nztm.geometry.distance(pt_nztm.geometry[0]) < 500]

A native spatial-filter API is on the eolas roadmap — for now, large region pulls + local filtering work.

Property research: parcels + titles + addresses

parcels   = client.linz("nz_primary_parcels", as_sf=True)
titles    = client.linz("nz_property_titles")
addresses = client.linz("nz_addresses", as_sf=True)

# Join titles to parcels via title_no
parcels_titled = parcels.merge(titles, on="title_no", how="left")

# Then spatial-join the addresses
parcels_full = parcels_titled.sjoin(addresses, how="left", predicate="contains")

This is the workflow behind property-search tools (PropertyGuru-style sites) — you're materialising what most NZ property platforms build on top of.

Pipeline use

LINZ datasets are served as full snapshots. Use sync_bulk / eolas_sync_bulk to keep a local file fresh — it issues a lightweight HEAD check and only re-downloads when the snapshot has changed.

from eolas_data import Client

client = Client("your_eolas_key")

# First call: full 1.6 GB download
result = client.sync_bulk("nz_parcels", path="/data/nz_parcels.parquet")
print(result.status)           # "downloaded"
print(result.bytes_downloaded) # ~1_650_000_000

# Next call: no-op when the snapshot is unchanged
result = client.sync_bulk("nz_parcels", path="/data/nz_parcels.parquet")
print(result.status)  # "unchanged" or "updated"

import pandas as pd
gdf = pd.read_parquet("/data/nz_parcels.parquet")
current = gdf[gdf["_eolas_is_current"] == True]

library(eolas)

result <- eolas_sync_bulk("nz_parcels", path = "/data/nz_parcels.parquet")
result$status  # "downloaded" first time, "unchanged" or "updated" after

library(arrow)
df <- arrow::read_parquet("/data/nz_parcels.parquet")
current <- df[df$`_eolas_is_current`, ]

# First sync
eolas sync nz_parcels --out /data/nz_parcels.parquet
# → downloaded (1.6 GB)

# Weekly cron — only re-downloads when snapshot changes
eolas sync nz_parcels --out /data/nz_parcels.parquet
# → unchanged (snapshot 5503437…)

See the Bulk downloads guide for format options, freshness levels, and scheduling recipes.

Source-specific notes

  • SCD2 for cadastral: parcels, titles, and addresses use SCD2 replication — each refresh adds a new version of any changed rows and marks the old one as expired. The _eolas_is_current, _eolas_valid_from, and _eolas_valid_to columns let you reconstruct the state at any past timestamp. Most users want WHERE _eolas_is_current = true to get the current cadastral state.
  • Size: nz_parcels (~3M) and nz_addresses (~3M) exceed the Free tier's 50,000-row cap. For the full extract you need Pro (limit=0, or the uncapped dashboard download) or the Enterprise Snowflake share. On Free, narrow the query (e.g. by region — a spatial filter is on the roadmap) or accept the 50,000-row sample. (Free monthly bulk-snapshot file downloads are on the roadmap — not yet available; the dashboard "Download" is a live query and applies the same Free 50k cap.)
  • Koordinates / WFS origin: most LINZ layers come from the data.linz.govt.nz Koordinates portal via WFS. Some layers (Landonline cadastral working data) come from the LINZ Exports API and are loaded via a different code path.
  • Earthquake-displaced marks: the canterbury_eq_* datasets capture geodetic marks that moved during the 2010-2011 + 2016 Canterbury quakes — important for surveyors reconciling pre- and post-quake datums.
  • Antarctic + Ross Dependency: NZ has Antarctic geodetic responsibility under the Antarctic Treaty; the antarctic_geodetic_* datasets are NZ-government data, not third-party.

Where to find more