Skip to content

Council data via eolas

New Zealand has 67 territorial authorities (TAs) — 11 city councils, 50 district councils, and 6 unitary authorities (which combine TA + regional council functions) — plus 11 regional councils on top. eolas aggregates ~600 datasets across 15 council clusters spanning the entire country, normalising the patchwork of ArcGIS, Koordinates, and bespoke portals into a consistent schema.

Coverage is ~99% of NZ population — 64 of 67 TAs are included, plus all 11 regional councils. (Three small TAs don't publish open data; we don't fabricate it.)

If you're doing planning, hazard, or infrastructure analysis where district-plan boundaries, flood hazards, mana whenua sites, or municipal assets matter — this is where to look.

How councils are clustered

We group councils into 15 regional clusters for two reasons:

  1. Regional councils share data with their territorial authorities. ECan publishes much of its work on the same Koordinates portal as Selwyn, Christchurch, Waimakariri, etc. — sharing a cluster means one pipeline.
  2. The clusters match how planners and infrastructure people actually work. "Wellington data" includes GWRC + Hutt + Porirua + Wellington City + Kāpiti + Carterton + Masterton + South Wairarapa + Upper Hutt — all relevant to most Wellington-region questions.
Cluster Datasets Population Guide
Auckland Council 20 1.7M District Plan overlays
Auckland Transport 10 1.7M Bus/ferry/train + cycle
Wellington Region Councils 76 540k GWRC + 8 TAs
ECan / Canterbury 85 670k ECan + 7 TAs (incl. Christchurch)
Co-Lab Waikato 79 530k 10-council consortium
Bay of Plenty 46 350k BoPRC + 5 TAs
Otago Councils 41 250k ORC + 5 TAs (incl. Queenstown)
Manawatū-Whanganui 46 260k Horizons + 7 TAs
Gisborne / Top of South 45 270k Gisborne, Marlborough, Nelson, Tasman
Southland Councils 48 105k ES + Invercargill, Gore, Southland District
Hawke's Bay Councils 33 180k HBRC + Central Hawke's Bay (other TAs in separate clusters)
Taranaki Councils 33 130k TRC + New Plymouth, South Taranaki, Stratford
Northland Councils 28 200k NRC + Far North, Kaipara, Whangārei
West Coast (Te Tai o Poutini) 24 33k WCRC + Buller, Grey, Westland (shared TTPP plan)
Napier + Whanganui 20 75k + 50k Napier City + Whanganui District (city-council layers)

What you'll find in council data

Most clusters share a similar palette of layers — what differs is the regional emphasis (e.g. earthquake hazards loom larger in Canterbury; cyclone exposure in Northland and Hawke's Bay).

District plan layers

District plans are the legal planning framework for each TA. Most TAs publish:

  • Zones — residential, business, rural, industrial, mixed-use, etc.
  • Designations — sites reserved for specific purposes (schools, transmission corridors, hospitals).
  • Overlays — heritage areas, coastal areas, natural features, hazards.
  • Precincts — special-character or themed sub-areas.

These come from each council's operative + proposed district plan; eolas serves the current operative version by default, with proposed versions tagged where available.

Hazards

A common set across most regions:

  • Flood hazards — historic flood extents, modelled future scenarios (typically 50yr / 100yr / 200yr / 500yr return periods).
  • Earthquake — fault avoidance zones, liquefaction risk, ground shaking amplification.
  • Coastal — erosion zones, coastal flooding, sea-level-rise inundation.
  • Tsunami — evacuation zones, modelled inundation (varies in return period from 100yr to 2500yr).
  • Volcanic — Taranaki + Auckland clusters have specific volcanic-hazard layers.

Cultural + heritage

  • Sites of significance to mana whenua / wāhi tapu — confidential locations marked where iwi/hapū have agreed to publication; many councils suppress exact locations for protection.
  • Heritage areas + heritage buildings — listed under the Resource Management Act and/or Historic Places Trust.
  • Archaeological sites — usually as point or polygon overlays.

Infrastructure + assets (varies)

  • Three Waters (water, wastewater, stormwater) — published by some councils; many keep it internal.
  • Roads — usually just centrelines or "council-owned" subset; the NZTA layer (Waka Kotahi guide) covers state highways.
  • Parks, reserves, council-owned buildings — varies in detail.

Environmental (regional councils only)

  • Coastal marine area — RMA boundary.
  • Air sheds — for monitoring + regulation.
  • Significant natural areas (SNAs) — biodiversity hotspots.
  • Resource consents — current discharges, takes, abstractions.

Refresh schedule

Weekly, Wednesday morning NZ time. Councils don't publish on a coordinated cadence — most update on their own irregular schedule when a plan change is notified, a hazard map is revised, or a new consent is issued. The weekly refresh catches whatever's changed at source.

meta = client.info("akc_aircraft_noise_overlay")
meta["last_refreshed_at"]
meta["source_last_modified_at"]

License

The vast majority of council open-data layers are CC-BY 4.0. A few specific layers have local restrictions (typically Three Waters infrastructure and cultural sites). Always check the per-dataset metadata before commercial redistribution.

Recommended attribution: "Source: [Council name], served via eolas (eolas.fyi). CC-BY 4.0." — substitute the specific council, since each is the licence-holder for its own data.

For multi-council analysis (e.g. nationwide hazard maps), use a generic attribution: "Source: NZ Councils via eolas (eolas.fyi). CC-BY 4.0 — see eolas dataset metadata for per-source attribution."

Common patterns

Find every council's district-plan zones

import pandas as pd

# Most clusters use a `*_district_plan_zones` or `*_dp_zones` naming pattern
datasets = client.list()
zones = [d["name"] for d in datasets if "district_plan_zones" in d["name"].lower() or "_dp_zones" in d["name"].lower()]
print(f"Found {len(zones)} district-plan zone datasets")

Hazard exposure for a parcel

The workflow most consulting / planning use cases follow:

import geopandas as gpd
from shapely.geometry import Point

# 1. Pick a location (lat, lon)
pt = gpd.GeoSeries([Point(174.7762, -41.2865)], crs="EPSG:4326")

# 2. Which TA contains this point?
ta = client.linz("nz_territorial_authority_2023", as_sf=True)
location_ta = gpd.sjoin(pt.to_frame("geom").set_geometry("geom"), ta, predicate="within")
ta_name = location_ta["ta_name"].iloc[0]

# 3. Load that TA's flood + earthquake hazards
# (dataset name varies; this is a Wellington example)
flood = client.wellington("gwrc_flood_hazard_extents", as_sf=True)
flood_at_point = gpd.sjoin(pt.to_frame("geom").set_geometry("geom"), flood, predicate="within")

print(f"Location: {ta_name}")
print(f"In flood-hazard area: {len(flood_at_point) > 0}")

A native "what's at this point" API is on the eolas roadmap. For now, the workflow is: fetch the relevant council layer + spatial-join in-memory.

District-plan compatibility across regions

District plans share a common vocabulary (zones, designations, overlays) but each council names columns differently. The eolas pipeline standardises basic columns (geometry_wkt, _eolas_* metadata) but per-council attribute schemas are preserved. Use client.info(name) to inspect the schema before joining cross-region.

Source-specific notes

  • Coverage gaps: 3 TAs (Chatham Islands, Stewart Island, Mackenzie District) don't publish open data and aren't in eolas. The council coverage memo tracks exactly what's in and out.
  • District plan version: most clusters serve the operative district plan. Where a proposed version exists, it's served as a separate dataset (suffixed _proposed).
  • Co-Lab Waikato is a consortium: 10 councils share one Koordinates portal — that's why their cluster has 79 datasets despite covering "smaller" individual councils.
  • Wellington Region's 76 includes Carterton, Hutt, Kāpiti, Masterton, Porirua, South Wairarapa, Upper Hutt, Wellington City + the regional council: it's effectively the entire Wellington administrative area.
  • Northland Councils' 28 includes Far North + Kaipara + Whangārei + NRC: which covers the entire Northland region.
  • ECan / Canterbury's 85 includes Ashburton, Christchurch, Hurunui, Selwyn, Timaru, Waimakariri, Waimate + ECan: the Canterbury Regional Council area minus the small Mackenzie District.
  • Hazard scenarios are modelled, not measured: a "100-year flood" layer is a probability-based hazard envelope, not a flood map of an actual event. For historical events, check *_historic_flood_extents or news archives.

Where to find more