Skip to main content

Last reviewed: 25 August 2026

Pā and archaeological landscapes

A pā is not a dot on a heritage layer.

The visible earthworks may be one part of a much wider landscape involving kāinga, cultivations, water, access routes, defensive sightlines, mahinga kai, nearby pā and later occupation.

GIS can help reconnect those relationships. It can also create false certainty if an archaeological point is treated as the whole place.

Start with landscape, not the site marker​

When reviewing a pā or archaeological place, useful context includes:

  • ridgelines and slope
  • freshwater access
  • coastlines and waterways
  • routes through the landscape
  • nearby cultivable land
  • known or possible kāinga
  • other pā and archaeological records
  • historical place names
  • old and current imagery

This can reveal patterns that are invisible when the site is viewed only as a point in a database.

Archaeological records answer a particular question​

ArchSite, maintained by the New Zealand Archaeological Association, records archaeological sites. Heritage New Zealand Pouhere Taonga also provides archaeology guidance and administers the archaeological authority process.

These systems are important sources. They do not provide a complete map of all Māori places of significance.

An archaeological record may be:

  • incomplete
  • based on a particular survey
  • spatially approximate
  • focused on material evidence
  • updated as new information becomes available

A place can be culturally important without being recorded archaeologically. Likewise, an archaeological point does not carry all cultural meaning associated with the place.

Heritage New Zealand explains that archaeological sites are protected under the Heritage New Zealand Pouhere Taonga Act 2014 and that an archaeological authority may be required for work that modifies or destroys a site.

Absence from a GIS layer is therefore not evidence that no archaeological site exists. Earthworks and development may involve legal processes that sit outside the GIS itself.

LiDAR can reveal form, not meaning​

Bare-earth LiDAR-derived terrain can make subtle earthworks visible beneath vegetation or difficult lighting.

Useful products include:

  • hillshade
  • multidirectional hillshade
  • slope
  • local relief models
  • contours

These can help identify terraces, ditches, banks, pits, tracks and other landform patterns.

A shape that resembles a terrace is not automatically a pā feature. LiDAR is a source for investigation, while interpretation can draw on archaeology, historical evidence and local knowledge.

See LiDAR for Māori GIS.

Compare several visualisations​

One hillshade direction can hide features aligned with the light source.

A useful desktop workflow is to compare:

  • hillshade from several directions
  • multidirectional hillshade
  • slope
  • contours
  • current imagery
  • historical imagery where available

The aim is to keep the visual interpretation proportionate to what the elevation data actually supports.

Historic imagery can show later change​

Aerial photographs can help identify:

  • vegetation change
  • earthworks before later damage
  • tracks and access
  • cultivation patterns
  • forestry
  • quarrying
  • erosion
  • urban expansion

The imagery date matters. A 1950 photograph does not show a pre-contact landscape. It shows one later moment that may preserve or obscure earlier features.

Build relationships between places​

Instead of one site layer, a research project may use several linked representations:

  • recorded archaeological sites
  • terrain features under investigation
  • known pā names
  • historic kāinga
  • water sources
  • routes
  • mahinga kai context
  • source documents

Keeping statuses explicit helps prevent a possible feature quietly becoming a confirmed site.

For example:

recorded_site
locally_identified_place
possible_terrain_feature
historical_reference
needs_field_review

Visibility analysis needs context​

GIS can calculate viewsheds from pā locations or ridgelines. This can be useful for exploring intervisibility, coast or river views, and possible communication relationships.

A viewshed is a model based on:

  • an elevation surface
  • observer height
  • target height
  • terrain resolution
  • assumptions about vegetation and buildings

It is evidence of modelled visibility, not proof of historical purpose.

Sensitive locations​

Precise archaeological and cultural locations can create risks including:

  • fossicking
  • theft
  • damage
  • unwanted visitation
  • exposure of wāhi tapu or urupā
  • cumulative inference when combined with other layers

A public educational map often works just as well with broad areas, generalised locations, landscape-level maps or selected public records, while detailed working layers can retain greater precision where that serves the kaupapa.

See Publishing maps with sensitive locations.

Field work​

A coordinate on a desktop map does not say anything about physical access to the land. Field work can involve ordinary land access arrangements, local tikanga, health and safety, and the archaeological requirements that apply to the activity.

It is also worth deciding what photographs, GPS observations or new records are useful to bring back into the project and who the resulting material is for.

A practical research workflow​

  1. Define the kaupapa and area.
  2. Gather known archaeological, historical and local source references.
  3. Review terrain, imagery and hydrology at landscape scale.
  4. Separate recorded sites from possible features.
  5. Link each interpretation to its source.
  6. Compare names and local context with the sources and people relevant to the kaupapa.
  7. Bring in archaeological expertise where the question needs it.
  8. Plan field work around the actual access, tikanga and legal context of the place.
  9. Keep working and shared versions separate where they need different levels of detail.
  10. Publish the landscape context and spatial detail that fit the purpose of the output.

Sources​

Last updated on