Guide

Active faults and property due diligence in New Zealand

An active fault finding is evidence about a mapped geological feature. It is not a forecast, not a verdict on the house, and not the same question as how hard the ground will shake. Knowing which of those it is changes what you do about it.

Part of How to research a property before buying in New Zealand.

What the finding actually is

When PropertyKete reports an active fault near a property, it is telling you one narrow thing: somewhere within a set distance of this land, geologists have mapped the surface trace of a fault classified as active. That is a fact about a map of the ground. It is not a forecast, it is not a measurement taken at the property, and it is not a judgement about the house.

The word “active” does most of the misreading. In everyday use it sounds like “currently moving”. In New Zealand practice it is a classification with a threshold attached, and the threshold reaches far enough back that a fault can be correctly classified as active with no recorded movement in all of this country’s written history. The definitions below are the ones GNS Science and the national planning guideline use, with their sources attached rather than paraphrased from general knowledge.

Four things a fault finding is not

It is not a prediction. The finding records a mapped feature. Even a fault’s recurrence interval — the average gap between its surface ruptures — is described in the national guideline as a statistical measure that may not predict future movement.

It is not a site assessment. Distance to a mapped line is not the same as an engineer’s judgement about what the ground under a particular building would do.

It is not the shaking question. Fault rupture is the ground tearing at the surface. Shaking is what travels outward from an earthquake, and the guideline that governs fault rupture planning says plainly that shaking causes most earthquake damage — and that it is a separate subject from the one it covers.

And its absence is not the reverse of it. No fault mapped within our rule does not establish that a property is free of earthquake-related hazard. It establishes that one regional-scale dataset has nothing within 500 metres.

What “active fault” means here

The word is a technical classification with a defined threshold, not a description of how dangerous something is. These are the definitions GNS and the national planning guideline actually use.

An active fault in Aotearoa New Zealand is one that has ruptured the ground surface within the last 125,000 years. The Taupō Rift is the exception, where activity is defined over the last 25,000 years instead. “Active” is therefore a statement about the geological past, not a statement about what will happen next.

New Zealand Active Faults Database — definition of activity · GNS Science

A fault’s recurrence interval is the average time between surface ruptures on it, and the national guideline treats it as the best available measure of the hazard an active fault presents. The guideline is explicit that this is a statistical measure only and may not be an accurate predictor of future movement — it gives the White Creek Fault, with a recurrence interval of more than 20,000 years, rupturing in the 1929 Murchison earthquake.

MfE guideline, section 7.1 · Ministry for the Environment (prepared with GNS Science)

Recurrence intervals are grouped into six classes: Class I is 2,000 years or less; Class II above 2,000 up to 3,500 years; Class III above 3,500 up to 5,000; Class IV above 5,000 up to 10,000; Class V above 10,000 up to 20,000; and Class VI above 20,000 up to 125,000 years. A fault whose average recurrence interval is longer than 125,000 years is not considered active. Determining a recurrence interval takes detailed investigation, usually trenching.

MfE guideline, table 7.1 and section 7.1 · Ministry for the Environment (prepared with GNS Science)

The national guideline covers surface fault rupture only. It states that it does not deal with the other earthquake-related hazards — strong ground shaking, liquefaction, uplift, subsidence, landslide and tsunami — and separately notes that shaking is what causes most earthquake damage. Fault rupture and shaking are different hazards, assessed differently.

MfE guideline, section 1 and section 3 · Ministry for the Environment (prepared with GNS Science)

What the mapping holds, and at what resolution

PropertyKete reads one specific dataset. What that dataset was built to show — and the scale it was built to be read at — sets the outer limit of anything our check can tell you.

The New Zealand Active Faults Database holds the surface traces of onshore active faults, with attributes describing each trace’s name, sense of movement, displacement and recurrence interval. It is compiled and maintained by GNS Science.

New Zealand Active Faults Database — dataset description · GNS Science

GNS publishes the database in two forms. AF250 is generalised for portrayal and use at 1:250,000 — a regional scale — and GNS states it is not to be used at a scale finer than 1:100,000. A separate high-resolution dataset, mapped at 1:10,000 or better, exists for selected areas and is the one designed for cadastral, property-level use.

Active Faults Database web map — conditions of use · GNS Science

Locational accuracy within the database varies from trace to trace, because traces were captured by different methods — aerial photo interpretation, theodolite survey, differential GPS — at different scales and for different original purposes.

New Zealand Active Faults Database — data quality · GNS Science

The national guideline warns directly against reading 1:250,000 fault data at property scale. It says a map should only be interpreted at the scale it was compiled at, that enlarging a 1:250,000 map causes a fault to be inaccurately portrayed and its placement interpreted wrongly, and that a key mistake is thinking a fault intersects a particular property when it does not. For planning purposes it says faults should be mapped at a minimum scale of 1:10,000.

MfE guideline, sections 5.1–5.2 and figure 5.2 · Ministry for the Environment (prepared with GNS Science)

The guideline states that most of New Zealand’s major active faults have been identified and mapped, at least on small-scale maps. “Most” and “at least” are the operative words: the mapping is a record of what has been found and compiled, not a guarantee that everything is in it.

MfE guideline, section 3 · Ministry for the Environment (prepared with GNS Science)

How councils and engineers use fault information

Fault rupture is managed through planning rules and building decisions rather than through a single number. This is the framework the national guideline sets out, which is what a council is usually working from when a fault matters to a consent.

A fault avoidance zone is a buffer either side of the known fault trace or likely rupture zone. The national guideline recommends a minimum of 20 metres either side, because intense deformation and secondary ruptures commonly occur within that distance of the primary rupture plane. A detailed fault study can justify a narrower zone; complex or distributed faulting can require a wider one.

MfE guideline, section 6.1 · Ministry for the Environment (prepared with GNS Science)

Faults are classified by complexity: well defined, where the trace is metres to tens of metres wide; distributed, where deformation spreads over tens to hundreds of metres and usually involves multiple traces or folds; and uncertain, where the trace has not been mapped in detail or cannot be identified. One fault can show all three along its length.

MfE guideline, table 8.1 · Ministry for the Environment (prepared with GNS Science)

What is acceptable inside a fault avoidance zone depends on the building as well as the fault. Building Importance Categories, running from 1 for structures presenting a low degree of hazard to life up to 4 for the most critical, are combined with the fault’s recurrence interval class to decide what development a council will allow. A house and a hospital are not treated the same way on the same ground.

MfE guideline, section 9.1 and tables 11.1–11.2 · Ministry for the Environment (prepared with GNS Science)

The guideline names the professionals equipped to locate and assess an active fault at a site: geologists with particular experience of mapping faults, and engineers with recognised qualifications and experience in geotechnical engineering.

MfE guideline, section 5.1 · Ministry for the Environment (prepared with GNS Science)

GNS publishes Fault Avoidance Zones and Fault Awareness Areas as their own map layers, separate from the fault traces themselves. A distance to a mapped trace is not the same object as a council’s or GNS’s zone, and the two should not be read as interchangeable.

GNS active fault map service — layer list · GNS Science

Exactly how PropertyKete checks it

PropertyKete is designed to help with early property research — the stage where you are working out what deserves a closer look. The active fault check is the clearest example of what that means: it is the one hazard check that runs everywhere in the country, and what it can and cannot resolve is unusually easy to state.

Every statement below describes the code that ships, not an intention. Where the behaviour looks blunt, it is — naming it precisely is the only way you can judge what the result is worth.

The limitation that matters most, in the guideline’s own terms

The dataset we read is published for use at 1:250,000. The national planning guideline says a map should only be interpreted at the scale it was compiled at, and that reading 1:250,000 fault data at property scale leads to exactly one mistake in particular — concluding that a fault intersects a property when it does not.

Our check computes a distance in metres and deliberately does not show it: a number on the page would imply we know where the fault is to the metre. It reports only what the dataset supports — whether a mapped trace lies within 500 metres — alongside GNS’s own accuracy class for that trace, which is not Accurate for roughly 55% of traces nationally.

GNS publishes a higher-resolution dataset for selected parts of the country, mapped at the scale planning work needs. PropertyKete does not hold it. Where a fault matters to a decision, GNS’s own web map and the council are where the better mapping is.

What the check does

The check runs for every New Zealand address. It is not limited to a council area, and it is the one hazard check that produces a finding nationwide.
It reads GNS Science’s active faults dataset at the 1:250,000 generalised scale (NZAFD-AF250). It does not read the high-resolution dataset GNS publishes for selected areas.
It finds the nearest mapped fault trace — the single closest one, not every fault in the area — and reports whether it lies within 500 metres. It does not report a distance in metres: AF250 cannot carry that precision, so the computed value is kept for our own reproducibility and never shown.
Where we have resolved a parcel boundary for the address, the measurement is taken from the parcel. Where we have not — commonly unit titles and rural addresses — it is taken from the address point instead, which will differ on a large section.
A fault mapped within 500 metres produces a finding, and that is the whole of it — one result, not a graded one. A trace at 10 metres and a trace at 480 metres read identically, because at this dataset’s mapping scale the difference between them is smaller than the uncertainty in where the line sits.
Beyond 500 metres the finding records that no active fault is mapped within 500 metres, and names the nearest one where it is within 5 kilometres, as context rather than as a concern. That one is described to the nearest half-kilometre, not in metres.
If the dataset is not loaded, the check reports itself as unavailable rather than returning a clear result. That guard exists so an empty database can never read as an absent fault.
The finding names GNS Science and the dataset it came from, and it is shown in the free preview on every report. It is never held back for a paid report or a Pro subscription.

Where the active fault check returns a result, city by city
City	Status	What that means
Auckland	Returns a result	Runs for every New Zealand address and is always shown in the free preview. It reports whether a mapped active fault trace lies within 500 metres, from GNS Science mapping compiled at 1:250,000 — a national screening scale, not a site survey. It does not give a distance in metres, because the dataset is not precise enough to place a fault within an individual property.
Wellington	Returns a result	Runs for every New Zealand address and is always shown in the free preview. It reports whether a mapped active fault trace lies within 500 metres, from GNS Science mapping compiled at 1:250,000 — a national screening scale, not a site survey. It does not give a distance in metres, because the dataset is not precise enough to place a fault within an individual property.
Christchurch	Returns a result	Runs for every New Zealand address and is always shown in the free preview. It reports whether a mapped active fault trace lies within 500 metres, from GNS Science mapping compiled at 1:250,000 — a national screening scale, not a site survey. It does not give a distance in metres, because the dataset is not precise enough to place a fault within an individual property.

Source: GNS Science, New Zealand Active Faults Database (NZAFD-AF250). The finding is shown in the free preview on every report and is never held behind a paid report or a Pro subscription. See what we check, city by city.

Reading the three results

Every PropertyKete check ends in one of three states, and they are not variations on a theme. Two of them look reassuring and only one is evidence.

The distinction that catches people is between the second and the third. “Checked, and none mapped” means a search happened and came back empty — a real result, bounded by what the dataset holds. “No result” means no search happened. Letting those two blur together converts a gap in our data into a statement about the property.

A fault is mapped near this property

A mapped active fault trace lies within 500 metres of the parcel. The finding names the fault where the dataset names it, and reports the accuracy class GNS assigns to that trace, but it does not give a distance — the dataset cannot support one. This is a prompt to find out what it means at this site.

What it does not settle: It does not establish that the trace crosses the property, that an earthquake will occur, or that the buildings are affected. At AF250’s mapping scale the position of the line is approximate, and the property-scale question is a site question.

Checked, and none mapped within the rule

The dataset was searched and no mapped active fault trace falls within 500 metres of this property. That is a genuine finding and it is worth having — it is different from having no information at all.

What it does not settle: It is a statement about what is mapped in one regional-scale dataset, not about the ground. Faults not in the database, faults below the mapping resolution, and every other earthquake-related hazard — shaking above all — remain entirely unaddressed by it.

No result

The check did not produce a finding, because the dataset was not available to it. The report says so explicitly rather than showing a clean result.

What it does not settle: Nothing has been searched and nothing has been ruled out. This state and the one above look similar on a page and mean completely different things, which is exactly why the report distinguishes them.

When a mapped fault actually changes something

Fault rupture is managed mainly through planning rules and building decisions. That is where a mapped fault turns from information into consequence — usually about what you can build, rather than whether you can buy.

Where it tends to bite

Building, extending or replacing a dwelling. Where a council has adopted a fault avoidance zone, the rules inside it depend on the fault and on what you propose.
Subdividing. Creating new lots on or near a mapped fault is where councils apply their strictest provisions, because it commits future owners too.
Anything of higher consequence than a house. The building importance framework treats a childcare centre, a workplace or a critical facility differently from a dwelling on the same ground.
A house already sitting on or across a trace. What was consented, and on what information, is a property-file question — and one an engineer may need to look at.
Insurance and lending conversations, which are not predictable from a distance measurement — an insurer prices its own view, and it is worth asking early.

What it does not do, in the ordinary case, is settle a purchase. Parts of several New Zealand cities sit within a few hundred metres of a mapped active fault, and houses are bought and sold there every week. The finding tells you a question exists; the answer comes from the council’s mapping and, where it matters, from someone standing on the site.

What to do when a fault is mapped near a property

The useful moves are mostly cheap, and run in an order that often ends the question before it reaches a professional.

Look at GNS Science’s own active faults web map for the address. It is free, it is the authoritative version of any fault line a third-party report shows you, and it carries the better mapping where that exists.
Check the council’s district plan and hazard maps. If a fault avoidance zone or fault awareness area has been adopted, that is the mapping the council will actually apply, and it may be far better resolved than ours.
Read the LIM. Natural hazards known to the council belong on it, along with the planning provisions and consent history attached to the land.
Ask the vendor, through the agent, whether any geotechnical or fault investigation has been done. Near a known trace, one often has been.
Get the property file where a building sits on or near the trace, to see what was consented and on what information.
Where you intend to build or subdivide, ask a planner what the fault provisions mean for that proposal — and a geotechnical engineer or engineering geologist to assess the site where the answer needs one.
Tell your solicitor what the finding is and what you plan to do with the property, so any conditions in the agreement reflect the real question rather than the map.

The sequence matters more than it looks. The free steps replace a regional-scale line with the mapping a council will actually apply, and where a professional is still needed, they let you brief one precisely instead of paying them to find what was already published.

The same finding, read on a real report

A clean result is evidence, not a guarantee

Every check came back clear. Here is what that is worth.

A report where the fault check ran and found nothing within 500 metres. Its commentary works through what that clean result is worth at a 1:250,000 mapping scale — which is the reading this whole guide is about.

Who can answer the questions a fault finding raises

Each entry says why a mapped fault in particular sends you to that person, and what to actually ask when it does. These are routes, not walls — on a fault question the council, a planner and an engineer routinely end up in the same conversation. For what each profession covers more generally, the full due-diligence guide sets them out side by side.

The council

Why them: A council is the one that decides whether a mapped fault constrains anything. Where it has adopted fault provisions, it holds mapping prepared for planning use — a far better resolution than the regional dataset our check reads — and it applies that mapping to consents on the land.

What to ask: Ask whether the property sits in a fault avoidance zone, fault awareness area or equivalent planning overlay, what mapping that is based on and at what scale, and what rules apply inside it. Order the LIM for the hazards recorded against the land and the consent history attached to it.

An engineer, including geotechnical

Why them: The gap between a regional-scale line on a map and this particular section is a site question, and the national guideline names geotechnical engineers and fault-mapping geologists as the people equipped to close it. Nothing in a screening report substitutes for someone assessing the ground.

What to ask: Ask a geotechnical engineer or engineering geologist to assess where the fault trace actually runs relative to the site, how well located it is, and what that means for the existing buildings or for what you intend to build. Where a fault study already exists for the area, ask them to interpret it for this property.

A planner or resource management specialist

Why them: Fault rules bite on a proposal, not on land in the abstract, and what is acceptable inside a fault avoidance zone depends on the building as well as the fault. Whether your extension, minor dwelling or subdivision is affected is a plan-rules question before it is an engineering one.

What to ask: Ask what the fault provisions mean for the exact thing you intend to build or create, whether it would be permitted or need consent, what building importance category it would fall into, and what geotechnical information the council would expect to see with an application.

A solicitor or conveyancer

Why them: A fault finding can affect what you are able to do with the land after settlement, which makes it a matter for the agreement rather than only for the plans. It is also where the consent history of anything already built on or near the trace gets tested.

What to ask: Tell them a fault is mapped near the property and what you intend to do with it, and ask what the offer should be conditional on — council confirmation of the fault provisions, a geotechnical report, or evidence that existing buildings were consented on the information the council held.

The agent

Why them: On a property near a known trace, a geotechnical or fault investigation has often already been commissioned, either by the vendor or for the subdivision the section came from. Asking is free and the report may answer the whole question.

What to ask: Ask whether the vendor holds any geotechnical report, fault investigation, engineer’s assessment or LIM for the property, and whether the council raised fault provisions during any consent the vendor obtained.

What a PropertyKete fault finding is not

A site-specific fault investigation: Our check measures distance to a mapped trace in a regional-scale dataset. Establishing where a fault runs across a particular section, and what that means for the ground under a building, takes investigation on the site.
The council’s fault mapping and rules: Where a council has adopted fault provisions, its own mapping is what applies to a consent, and it is generally better resolved than the dataset we read. We do not hold it and we do not reproduce it.
An earthquake hazard assessment: This check addresses surface fault rupture only. Ground shaking, liquefaction, landslide and tsunami are separate hazards, and PropertyKete produces no finding on any of them anywhere in New Zealand.
A LIM: The council’s formal record of what it holds about the property, including natural hazards known to it and the planning provisions attached to the land. We hold no equivalent of it.

Checklist

Checked whether the active fault check returned a finding, returned none within the rule, or produced no result at all.
Read the reported distance as an indication of proximity from a regional-scale dataset, not as a measurement of where the fault runs.
Viewed GNS Science’s own active faults web map for the address, including the higher-resolution mapping where it exists.
Checked the district plan and council hazard maps for a fault avoidance zone, fault awareness area or equivalent overlay.
Read the LIM for natural hazards recorded against the land and for planning provisions attached to it.
Asked the vendor whether any geotechnical or fault investigation has already been done.
Obtained the property file where a building sits on or near the mapped trace.
Engaged a geotechnical engineer or engineering geologist where the answer depends on the site rather than the map.
Confirmed with a planner what the fault provisions mean for any build, extension or subdivision being considered.
Kept fault rupture separate from shaking, liquefaction, landslide and tsunami, none of which this check addresses.
Told the solicitor what the finding is, so the agreement’s conditions reflect it.

A mapped active fault is one of the more misread things a property report can show you, and the misreading runs in both directions. It is not the disqualification it sounds like, and its absence is not the clearance people take it for. It is a mapped feature, recorded at a scale that was never meant to answer questions about one section, pointing at a question worth asking.

So ask it in the right order: the free mapping first, the council’s rules second, and a professional only where the answer genuinely depends on the site. And hold on to the difference between a check that searched and found nothing and a check that never ran — on this subject more than most, they are not the same sentence.

Sources

Every statement on this page about active faults, the database or planning practice was checked against these sources on the date shown. The fault data in a PropertyKete report comes from GNS Science.

New Zealand Active Faults Database — GNS Science. Checked 2026-09-05.
Active Faults Database web map — dataset scales and conditions of use — GNS Science. Checked 2026-09-05.
Langridge et al. (2016), “The New Zealand Active Faults Database”, New Zealand Journal of Geology and Geophysics 59: 86–96 — GNS Science (the citation GNS asks users of the NZAFD to give). Checked 2026-09-05.
Planning for Development of Land on or Close to Active Faults: a guideline to assist resource management planners in New Zealand — Ministry for the Environment (prepared with GNS Science). Checked 2026-09-05.

This guide is general information, not advice

It describes how active faults are mapped in New Zealand and how that mapping is used. It is not legal, engineering, insurance, valuation or financial advice, it does not take account of your circumstances or of any particular property, and it is not an assessment of earthquake hazard at any address. Every decision in it should be made with the relevant professional named above.

The full due-diligence guide · What a LIM covers · What we check, city by city · Data sources