At fault

Recent earthquakes in New Zealand and Japan have tragically demonstrated the immense devastation that can be caused by the planet's natural activity.

Geoscientists and engineers in New Zealand are at the forefront of assessing earthquake hazards and suggesting solutions to minimise the risks to people and infrastructure. Yet, despite, more than a century of scientific investigation, the underlying physical processes that control how earthquakes start, propagate and stop, remain elusive.

Now Department of Geology researchers are drilling deep into New Zealand's Alpine Fault to gain a better fundamental understanding of the processes involved and hopefully improve future earthquake risk management significantly. After successful shallow drilling in the Southern Alps last year, plans are being made to bore about 1.5 kilometres into the fault zone.

This fault was one of the attractions that lured Geology Head of Department Professor Dave Prior from the UK to Otago.

“New Zealand's Alpine Fault is the best place to study the physical processes that control earthquakes and Otago is the best place to work from,” says Prior, who began researching the zone in the 1980s.

The current Deep Fault Drilling Project is partly funded by the International Continental Scientific Drilling Programme, a global organisation trying to get a better understanding of the earth's crust.

The outermost skin of Earth – the crust and the upper mantle – is made up of several rigid tectonic plates.The boundary of the Australian and Pacific plates runs along the length of New Zealand, with the east side of the South Island (the Pacific plate) moving south-west relative to the west side (the Australian plate).

The Southern Alps are the result of the Pacific plate also riding up over the Australian plate. Part of the boundary where the plates converge gives rise to the Alpine Fault, which stretches about 300 kilometres along the spine of the Alps, and has produced several large earthquakes over the last 1,000 years.

“The geology gives us the history,” says Prior. “We have a record in the rocks and that record tells us that we are late in the cycle that builds up to an earthquake.

“We are confident that the fault is in a highly stressed state and there will be a big quake – but we don't know when. It could be tomorrow or in 10, 20 or 200 years.”

The department is combining national and international co-operation with new research tools to try to find out more.

“As we learn more about the signals that tell us what state a fault is in, we might be able to make better assessments of how close to a state of rupture a fault is.”

The Alpine Fault is geologically unusual in two ways, making it both easier to study and a prime candidate for drilling. Firstly, the active zone of earthquake initiation and rupture is relatively shallow – roughly less than 10 kilometres deep – and secondly, historical movements have exhumed rocks from all depths of the active zone to the surface.

“We've already collected examples of rocks that were active from the surface to as deep as 20 kilometres down, depths we are never going to access by drilling. But by going down as far as we can, we can compare the new material we find with what we have inferred to be going on,” says Prior.

“Drilling means we can make measurements of environmental conditions that we can't interpret from a rock found on the surface. When we've measured actual conditions, then we can verify or modify our models in a realistic way.”

Otago's Dr Virginia Toy is a co-principal investigator for the drilling project with Dr Rupert Sunderland of the Institute of Geological and Nuclear Sciences (GNS) and Professor John Townend of Victoria University, and collaborators from around New Zealand and the world are supporting the project.

Otago's analysis of rocks has received a boost with a new electron backscatter diffraction laboratory set up by Prior, who was an early pioneer in introducing the technique to earth sciences some 15 years ago.

“Rocks are aggregated crystals and the electron backscatter diffraction microscope works on these, mapping the distribution of crystals, which can tell us such things as the stress conditions in which a rock deforms.

“We can use electron microscopy to study rocks that have been deformed naturally and then apply the same techniques to rocks that we've squashed artificially in the laboratory and compare results.

“Fault zones are all about rocks deforming. In the top 10 kilometres rocks can break and slide past each other, rather like sand between two pieces of wood. Below 10 kilometres it's hotter and rocks can bend and change shape without breaking.

“The drilling programme will help us take a material-science approach to understanding rocks and how they deform, allowing us to use analytical tools to help us understand the processes by which rocks can change shape and fracture and slide past each other.”

Prior's own research into the properties of ice is analogous to the work with rocks, and may save time and effort.

“My experiments in ice may be only loosely related to the drilling programme, but they can explain a wide range of properties and problems. We can do more with ice than we could do if we tried to do the experiments with rocks, as ice is much easier to deal with.”

Comparisons with other well-researched fault lines can be useful to a point, says Prior.

“The deeper part of the Alpine Fault is almost certainly moving all the time – through processes such as dislocation creep and grain boundary sliding – but this is not reflected on the surface.

“In the San Andreas Fault, in California, you get obvious creep close to the surface, where zigzag breaks in fence-lines and roads show what is happening. But the majority of the fault surface in Westland shows no signs of any surface creep at all, which suggests that pressure is building up, because GPS monitoring shows that the underlying tectonic plates have a substantial rate of movement.

“As hotter rocks rise up from depth they create a steep geothermal gradient, resulting in a fault zone that is not so deep as the San Andreas Fault, for example.

“This easier warming results in thermally activated creep – so you don't have to go so deep to find the Alpine Fault zone. As a result, the range of earthquakes that occur is different. There's a reduction in the number of earthquakes, but they're sizeable when they do occur.”

The drilling programme intends to extract rocks from as deep as 1.5 kilometres and use geophysical tools to measure such things as temperature, pressure and seismic and electrical properties, as well as placing instrument packages down the boreholes to continue to monitor those measurements.

“We need to see if we can get a better idea of the precursor signals for a major event,” says Prior. “It's similar to trying to understand how volcanic eruptions work and how imminent they are, but the advantage with a volcano is that we know where it's going to happen and the signals are close to the surface.

“What we are doing is making sure the area is instrumented up so we can record what happens immediately before a major event. And, although we'll only have this information after the event, we'll then be in a better position to judge how well we can signal other similar events in the future.”

So what are the chances of picking when the next big one is coming?

“Even if we have a full understanding of the physics, there are elements that are chaotic,” says Prior. “My personal opinion is that we are never going to be able to predict earthquakes.”

– NIGEL ZEGA