The Alpine Fault: Is New Zealand Prepared?

By Jesse Dykstra 20/11/2013

The central section of the Alpine Fault. Source:
The central section of the Alpine Fault (Source:

This post is the first of two articles which explore the potential impacts of the next great earthquake on the Alpine Fault and consider how prepared New Zealand is for that event.  This builds on a previous post which describes the physical properties of the Alpine Fault in more detail.

Improved Understanding of Earthquake Hazards in New Zealand

Damage in Christchurch following the 22 Feb. 2011 earthquake
Damage in Christchurch following the 22 Feb. 2011 earthquake

One of the positive outcomes of the Canterbury earthquakes of 2010 and 2011 is an increased focus on improving our understanding of earthquake risk in other seismically active regions of New Zealand,  including the Wellington area, and the West Coast of the South Island, where our largest and most active fault, the Alpine Fault, is due to rupture.

The imminent rupture of the Alpine Fault is a subject which has recently been the beneficiary of increased research focus in the scientific community. Several recent studies have improved our understanding of the rupture history of the Alpine Fault, including significantly extending the record of great earthquakes (i.e. > Mw 8) out to about 8,000 years before present. This research, which has been published in the leading journal Science, has confirmed a recurrence interval of 330 years for great earthquakes on the Alpine Fault, and that the probability of such a rupture occurring within the next 50 years is about 30%.

Has Public Awareness About the Alpine Fault Been Skewed by Our Perceptions of the Canterbury Earthquakes?

Has this improvement in our understanding of the behavior of the Alpine Fault translated directly into enhanced public awareness of the risks associated with the next great earthquake? Well, perhaps not as much as scientists and risk analysts would like to believe. Certainly the average person in New Zealand is well aware of the devastation caused by the Canterbury earthquakes, as well as the recent damage and apprehension caused by the Cook Strait earthquakes (centered near Seddon), but I would argue that those events, while increasing our general awareness of the inevitability of a future great earthquake on the plate boundary, have actually skewed our overall perception of the risks associated with the next Alpine Fault Rupture.

I sense that many of us continue to have a sense of complacency about seismic hazards in New Zealand – we have “gotten through” the recent earthquakes, so we must have had the “get ready” part sorted, right? Having ridden out the Darfield (Mw 7.1) and Christchurch (Mw 6.3) earthquakes, and thousands of aftershocks, we have a pretty good idea of what mother nature can throw at us, right? After all, how much worse can an earthquake get than the virtual destruction of a city’s entire CBD?

In the aftermath of the 22 February 2011 earthquake in Christchurch
In the aftermath of the 22 Feb. 2011 earthquake in Christchurch

We must remember that while large parts of Christchurch were devastated by the recent earthquakes, one only has to drive an hour in any direction from the CBD to find areas that appear to be completely unaffected by the earthquakes; there simply isn’t any evidence of significant earthquake damage beyond about a 50 km radius from Christchurch City.

This will not be the case following a great (Mw 8) earthquake. An Alpine Fault earthquake will be a very different beast than the Canterbury earthquakes, with a resulting truly national-scale disaster of much greater impact than any historical earthquake in New Zealand.  The entire South Island will be affected in some way, with indirect, but profound impacts to the entire country.

Globally, large earthquakes have caused some of the worst natural disasters over the last 50 years, including:

  • 2011  Mw 9 Tōhoku earthquake and tsunami (Japan),  18,000 deaths
  • 2010 Mw 7.0 Haiti earthquake, ~ 200,000 deaths
  • 2008 Mw 7.9 Sichuan earthquake (China),  ~90,000 deaths
  • 2005 Mw 7.6 Pakistan earthquake,  ~100,000 deaths
  • 2004 Mw 9.3 Sumatra-Andaman earthquake & tsunami (Indian Ocean),  ~275,000 deaths
  • 1995 Mw 6.9 Great Hanshin earthquake (Kobe, Japan),  5,500 deaths
  • 1976 Mw 8  Tangshan  earthquake (China),   240,000 – 800,000 deaths

In addition to causing much loss of life, these earthquake disasters also had significant and far-reaching economic impacts. The Chinese government has estimated the cost to repair the damage caused by the 2008 Sichuan earthquake at about $150 Billion U.S. dollars (approximately the same as New Zealand’s GDP).

If the Alpine Fault were to rupture in the near future, the most severely-affected areas along the West Coast will be sparsely populated, so it is likely that the associated death toll will be much lower than that resulting from large earthquakes in heavily populated regions of the world. This is a benefit of living in an area with low population density; however, regardless of how many people lose their lives, we can be sure that there will be significant social and economic impacts following an Alpine Fault earthquake.

Regional, Rather than Local Impact

Length of Fault Rupture & Total Energy Released

In general, the severity of ground shaking during an earthquake is directly related to the area of rock within the fault slip zone – the larger the area which slips, the more energy is released, with stronger and longer-lasting ground shaking as a result. The main section of the Alpine Fault is up to 450 km long; when that length of fault ruptures we can expect to see horizontal displacements of up to over 8 m, and vertical displacements of up to over 4 m.

These displacements could tear open the earth along hundreds of kilometres of the fault trace, and approximately 1000 times more energy will be released during such an earthquake than was produced by the Mw 6.3 earthquake that devastated Christchurch’s CBD on 22 February, 2011. The initial ground shaking will likely last 2-3 minutes, with subsequent strong aftershocks (up to >Mw 7) going on for days and up to weeks after the main event.

Modeled Earthquake Shaking Intensities, with Alpine Fault clearly visible running along the spine of the Southern Alps. Source:
Modeled Earthquake Shaking Intensities, with Alpine Fault clearly visible running along the spine of the Southern Alps. Source:

The last Alpine Fault rupture, which generated an earthquake of magnitude 8.1, occurred in 1717. At least 380 km of the fault ruptured, from Milford Sound to the Haupiri River; although this predates written records in New Zealand, geological records preserve evidence of wide-spread landscape damage caused by this event.  These records include enhanced periods of sedimentation (due to increased landsliding on hillsides) or gravel aggradation in river valley bottoms which caused the burial and death of forests, inundation of river valleys caused by sudden changes in channel geometry.

Similar landscape changes occurred after the 2008 Sichuan earthquake (Mw 7.9), which involved a rupture length of about 240 km, and affected an area approximately 900 km long by 600 km wide (i.e. an area significantly larger than the South Island).

Communities at Risk: Ground Shaking & Liquefaction

The magnitude of a future great earthquake on the Alpine Fault is likely to be such that at least “strong” (MMI VI) ground shaking will be experienced in even the furthest reaches of the South Island, as well as the lower North Island. Close to the fault trace catastrophic shaking intensities (up to MMI XII) will occur (see table below for a description of shaking intensities).

Fortunately, there aren’t any major cities very near to the Alpine Fault, but many smaller communities from Springs Junction  through to Franz Josef and Fox Townships and as far south as Haast are all located very close to the fault trace, and will probably experience severe ground shaking during an Alpine Fault earthquake, at least as intense as that which damaged Christchurch’s CBD during the 22 February 2011 event.

A bit further away from the fault, communities such as Queenstown, Wanaka and Greymouth will likely experience significant and destructive ground shaking (up to MMI VIII). The largest South Island cities of Christchurch and Dunedin will likely experience ground shaking of up to MMI VII, similar to that felt in many parts of Christchurch during the September 2010 Darfield earthquake. That level of shaking caused significant damage to unreinforced masonry buildings in Christchurch, but most infrastructure damage was caused by liquefaction and lateral spreading over susceptible soils throughout the greater Christchurch region.

Moment Magnitude Scale of Earthquake Shaking Intensity

While the intensity of shaking felt in Christchurch during an Alpine Fault great earthquake may be similar to that felt during the Darfield earthquake, the duration will be much longer (2-3 minutes compared to 40 seconds), so extensive liquefaction should be expected again.

New construction in Christchurch should benefit from updated building standards and a much-improved understanding of the effects of ground shaking on liquefiable soils, but other communities which are relatively distant from the Alpine Fault (e.g. Invercargill, Dunedin, Timaru, Nelson, Blenheim) are also at least partially built on liquefiable soils. Assets built on susceptible soils will likely suffer some damage during prolonged moderate ground shaking, especially unreinforced masonry buildings and underground utilities.

Impact on Lifelines & Critical Infrastructure

In addition to causing widespread damage to land, buildings and underground utilities across much of the South Island, a future Alpine Fault earthquake will almost certainly sever the major lifelines that cross the fault, including State Highways 6 and 73, and the sole rail link to the West Coast, through Arthur’s Pass. Ground shaking during the initial event, and subsequent aftershocks, will trigger many large landslides, especially on steep hillsides within perhaps 100 km of the fault line.

GNS Science estimates that great earthquakes (i.e. >Mw 8) can trigger large landslides up to 300 km away from the fault rupture, and documented over 400 new landslides following the 2003 Fiordland earthquake (Mw 7.2) . Where lifelines cross the actual fault trace, ground offset and rupture will directly damage roads and rail lines; however, there are also a number of indirect ways that ground shaking can damage lifelines:

  • bridges damaged by intense ground shaking
  • bridges damaged by debris flows which travel down valleys following slope failures in mountainous catchments (especially on the West Coast, where there are many rivers crossing the narrow swath of gently sloping land between the mountains and the sea)
  • road and rail lifelines rendered impassable by landslides in steep or mountainous areas (e.g. near Haast Pass and Arthur’s Pass, as well as Lewis Pass and the Buller Gorge)
  • flooding damage to road/rail links due to river channel avulsion or landslide-dam-break in valley headwaters

    Landslides and bridge collapse following 2008 Mw 7.9 earthquake in Sichuan Province, China (source: University of Deleware)
    Landslides and bridge collapse following 2008 Mw 7.9 earthquake in Sichuan Province, China (source: University of Deleware)

Immediately following the next Alpine Fault rupture, isolated communities which rely on these lifelines will likely become effectively cut off from the rest of the country. West Coast communities may be particularly hard hit, especially given their proximity to both the Alpine Fault and the western slopes of the Southern Alps, the number of rivers/bridges between communties, and the limited number of transport routes, all of which travel through steep mountainous passes or river valleys. Depending on the time of day, and time of year when the earthquake occurs, thousands of people could become stranded on roads or rail lines which are impassible,  and some may be directly impacted by landslides, falling rock, road collapse, or bridge failure.

It will probably take several weeks to months to fully restore lifelines and access to the most isolated communities, during which time land-based transport will be very difficult, especially to the West Coast. Ongoing large aftershocks (up to >Mw 7) will continue for weeks and months after the intial earthqauke, meaning areas that could be affected by landslides or rockfall will not be safe to work in. Following the Sichuan eartquake, 158 relief workers were killed by landslides as they tried to repair roads in the weeks immediately following the main earthquake. How long after the main earthquake will it be before it is safe enough for roading crews to begin repair work? This is very difficult to predict, but eventually, many bridges will have to be at least temporarily repaired, and dozens (if not hundreds) of large landslides cleared through through the main passes.

Landslide blocking road following 2008 Mw 7.9 earthquake in Sichuan Province in western China (

The initial emergency response will require massive mobilisation of resources from communities that are less affected (e.g. Nelson, Christchurch, Dunedin, Invercargill, Wellington). Port facilities and airports may be severely damaged as well, so there is no guarantee that critical emergency supplies and fuel and equipment will be readily available, or be able to be transported to communities in need. Services that we normally take for granted, such as electricity, clean drinking water and telephone will likely be severely damaged, particularly close to the fault. While some communities may be able to quickly establish temporary drinking water supplies, restoration of electricity may take weeks, or even longer, as power lines will likely be extensively damaged by landslides, and some power stations and storage facilities will require damage assessment (at a minimum), to become operational.

Milford Sound and Displacement Waves

Poster from upcoming Norwegian film based on Tafjord Disaster of 1934 (source:
Poster from upcoming Norwegian film based on Tafjord Disaster of 1934 (source:

I believe that the potential impact of an Alpine Fault earthquake on the iconic tourist destination of Milford Sound deserves consideration – located within some of the most rugged mountain ranges of Fiordland, Milford Sound receives up to over 600,000 visitors per year. The seaward entrance of the 13 km-long fiord is crossed by the Alpine Fault, and there is ample geological evidence of many very large prehistoric landslides preserved both in the valley bottoms near the head of the fiord (where the Milford village, new visitor centre/cruise boat terminal, airport and docks are located), and on the fiord bottom itself. Most of these landslides were probably triggered by large earthquakes, and some of them would have fallen from the steep mountainsides high above the fiord, generating large displacement waves (tsunami) upon impact with the water.

If such an event were to occur during a peak tourist time at Milford Sound, several hundreds or thousands of people could be killed; landslide displacement waves up to 70 metres high have killed hundreds of people in the Norwegian Fjords during the last century. Other fiords and steep-sided glacial lakes of southern New Zealand are also susceptible to displacement waves caused by coseismic landslides (e.g. Doubtful Sound, Te Anau, Manapouri, Queenstown, Wanaka, Hawea).  This under-appreciated hazard in New Zealand is the subject of my recently-completed doctoral thesis, and I will cover some of the implications of that work in more detail in future posts.

Editor’s note: unfortunately this was the last post in Dr Dykstra’s Alpine Fault series.

0 Responses to “The Alpine Fault: Is New Zealand Prepared?”

  • Hi Jesse, very informative article, I am just wondering if you have written part 2 yet and if so where I can read it?

    Many thanks

    • Thanks for the reminder Shery – how time flies! I have most of the 2nd part written, so will publish it over the next week or so.

  • Good article.
    Whenever I read about the Alpine fault there is talk of Mag 8…u gotta think it is possible/probable Mag 9.
    Look fwd to part 2.

    Cheers David

    • Thanks David,

      My understanding is that the largest earthquakes (ie. those greater than about M8-8.5) are generally those on major subduction zone plate boundaries. These “megathrust” earthquakes release huge amounts of energy because the surface area of rock that is locked together is so large. In contrast, the Alpine Fault is a strike-slip fault, where two plates are predominantly sliding past one another. Due to the limited length and surface area of the fault, it is unlikely that it could produce a M9 eathquake – however, be assured that a M8 event on the Alpine Fault is not something to be taken lightly.

  • An 8.5 on the alpine to then trigger a 9.0 on the wellington fault is not something to look forward to.