Ask an Expert: Did dairy farming trigger quakes?

01:07, Sep 02 2011

The Canterbury Plains, fed by alluvial water, had a concerning deficit of water flow during 2010. Does nature's system of copious water keep the Plains underground balanced and settled?

Did the lack of water, due to the many dairy farms taking it on its way to the sea and aggravated even more by a dry year, affect the magnetic plate's temperature and set off the Greendale Fault, which, in turn, triggered the other faults to wake up? Not my theory; I heard it from someone else but it made me curious enough to question it. - HEATHER

To answer this question, I need to consider the structure of the Earth's crust. Science suggests the lower crust behaves ductily through thermally controlled recrystallisation of minerals, the upper crust is brittle, with fracturing controlled by rock friction, and it is strongest and most resistant to deformation somewhere in between the two processes at mid-depths.

In Canterbury, the crust is about 28 kilometres thick and the Greendale Fault initiated its rupture about 11km below the surface, probably in greywacke rocks that contain very little (less than 1 per cent), if any, water. The overlying porous gravels of the Canterbury Plains are just a thin veneer (about 300 to 600 metres thick). The water table in these gravels fluctuates, caused by variations in natural recharge (rain, snow, etc) and human extraction, but the largest subsurface water-table fluctuations are small (tens of metres) compared with the thickness of the crust.

The thermal influence of such water fluctuations is tiny, unlikely to be transmitted below the gravel aquifers, and negligible compared with the general increase in temperature towards the centre of the Earth.

It is well-known that small earthquakes can be induced by pumping water down boreholes or filling lakes for hydro-electric storage dams, so my understanding is that pressure (rather than temperature) changes have greater potential to influence crustal strength and induce seismicity. At the mid-crustal site where the Greendale Fault rupture began, the influence of lowering the near-surface water table, even if it could be transmitted to this site, cannot be more than the weight of the water, which is far smaller than the weight of about 11km of overlying rock.


The lower water table in 2010, however, would decrease fluid pressure, increase the relative friction on brittle faults and fractures, thereby making it harder for small faults and fractures to slip and cause quakes. So for this reason, I do not believe fluctuations in the surface aquifers could have been the "straw that broke the camel's back". If anything, they would have taken a tiny amount of the pressure off. - DR SIMON COX

Geologist, GNS Science.

It has generally been stated that the South Island is divided geologically in that the West Coast moves north 5 centimetres a year and everything east of the Main Divide moves south 5cm a year. The fact that Christchurch has had earthquakes in the past suggests to me that Banks Peninsula does not want to move at the same rate as the rest of the eastern South Island. The first major quake happened inland at Greendale. The slip was directly lined up with the north edge of Banks Peninsula.

I consider this part of Canterbury, after being held up by the peninsula in the general move south, finally moved, thus loading up pressure on the remaining land approaching the peninsula from the north. Then the February quake happened. This would explain why Lyttelton is now nearer Christchurch, and the Port Hills are higher. Unfortunately, because of the general movement of the Canterbury Plains coming up against the Port Hills, quakes will probably happen again. What are your thoughts? - PETER McCULLY

It is a good question to ask what role the extinct volcanoes of Banks Peninsula have played in the Canterbury earthquake sequence. Before trying to answer that question, I'll go back a couple of steps.

The ongoing tectonic-plate motion means that in the South Island the Pacific Plate is moving about 40 millimetres each year in a southwesterly direction relative to the Australian Plate. So the east coast of the South Island is moving about 40mm a year southwest relative to the West Coast. Looked at the other way, the West Coast is moving about 40mm a year northeast relative to the east coast.

Viewed from the Main Divide, the West Coast is moving about 20mm a year to the northeast and the east coast is moving about 20mm a year to the southwest.

The boundary between two tectonic plates rarely consists of just a single geological fault, with everything on one side moving one way and everything on the other side moving the opposite way. Instead, there is a plate boundary deformation zone where many geological faults are active. An active geological fault is one that has had quakes in the past 10,000 years or so.

In the case of the South Island, the whole region between the east and west coasts, and extending offshore on both sides, is part of the plate boundary deformation zone. The Alpine Fault, which runs along the base of the Southern Alps on the West Coast, is the biggest of these faults. It has large quakes every few hundred years on average, in which the two sides of the fault slip past each other by four to eight metres. The Alpine Fault accounts for about 30mm a year of the 40mm a year of plate motion.

The other 10mm a year of plate motion causes stressing, straining and deformation of the Earth's crust on either side of the Alpine Fault, extending beyond both the east and west coasts. When the strain in a particular region builds up beyond the breaking strength of the rocks in the crust, the crust will fracture. The rocks on the two sides of a geological fault slip past each other and the energy released by the movement of the rocks causes an earthquake.

We therefore expect that quakes can happen anywhere in the plate boundary deformation zone. But the further away from the Alpine Fault, the rarer we expect the quakes to be. This was the case for the Canterbury quake on September 4. It is probable that the previous quake on this fault occurred many thousands of years ago.

Once the September quake happened, it caused changes of stress in the region around the quake, which resulted in other rocks in the region reaching failure and thus causing other quakes. This is the aftershock sequence following the original quake.

If a region of the Earth's crust is made up of rock of a uniform type, then when it is put under stress it will deform, or strain, in a uniform way. But if the crust is made up of varying rock types, the stress may become concentrated at the boundaries between the different rock types. And now we get back to the role of the Banks Peninsula volcanic rocks. The Banks Peninsula volcanoes erupted through the Earth's crust between about six million and nine million years ago. The volcanoes are now extinct, but they have left a whole region of different, volcanic rock emplaced within the pre- existing greywacke rock. One of these volcanoes is the Lyttelton Volcano centred on Lyttelton Harbour. When the volcanoes were erupting millions of years ago they would have caused earthquakes to occur around the edges of the volcanoes. These earthquakes may have occurred on pre-existing geological faults, or new faults may have been created.

After the September quake, it is likely that stress concentrations occurred around the edges of the volcanic rocks on western Banks Peninsula. In turn, these stress concentrations might have influenced the locations of the February 22 and June 13 quakes. It is even possible that these quakes occurred on faults that were also active when the volcanoes were being created more than six million years ago. Scientists are working on these questions.

Going back to the reader's questions, I do not think the Canterbury quakes occurred where they did because Banks Peninsula is reluctant to move with the rest of the Earth's crust. But I do think that the contrasting rock types between the Banks Peninsula volcanics and the surrounding greywacke rocks may have resulted in stress concentrations that influenced the locations of the quakes. And as for the question of whether there may be future quakes in the region, this certainly remains a possibility. The best estimates of the likelihood of future quakes in Canterbury have been calculated by seismologists at GNS Science and can be seen at http:/ / canterbury-quakes/ aftershocks/, and this site is updated regularly. - DR JOHN BEAVAN,

Geophysicist, GNS Science.

* Email your questions to

The Press