The where and when

of New Zealand

earthquakes

by Lee Aitken, Terry Webb

Clever science and technology that accurately measure and locate earthquakes are too recent for our scientists to have detailed pattern books for earthquake behaviours in all parts of New Zealand.  On top of all this, our written history is too short for seismologists to sort out patterns of normally recurring earthquake activity from activity that heralds an earthquake emergency.  Identifying the where and when of local earthquakes is absolutely fundamental to planning ways of lessening earthquake disasters in the future.

It is possible that different parts of New Zealand have such different earthquake environments that identifying patterns for some areas may be decades away.  And, of course, some areas may not be prone to repetitive earthquake paterns at all.

In the late 1980s, however, the national network of seismographs (instruments that record earthquake waves) was upgraded so that earthquakes could be located more accurately and efficiently.  This has confirmed some previous observations and interpretations about where earthquakes originate under different areas of New Zealand.  New patterns that had not been previously identified, however, have also shown up.  More research is needed to understand what causes these patterns of seismic activity, so that we can better estimate the future hazard posed by earthquakes.

The link between earthquakes and the tectonic-plate boundary that runs diagonally across the country is clearly demonstrated by plotting the depths of earthquakes.


Figure 1- Earthquakes under the North Island from Hawke's Bay to Kawhia.

One of the clearest pictures of the role of plate tectonics in New Zealand's earthquake history comes from plotting the depths of earthquakes along a line from Hawke's Bay to Kawhia (figure 1).  The earthquakes are relatively shallow in the east, but become progressively deeper towards the west.  This almost exactly follows the dip of the Pacific Plate as it dives under the Australian Plate - on which the North Island sits.  Earthquakes can occur all along the surface of the diving plate as it travels beneath the North Island into the interior of the planet.  Earthquakes more than 600 km deep have been recorded off the Taranaki Coast at which deptht he diving plate loses its rigidity and ability to fracture and create earthquakes.


Figure 2 - New Zealand's deep earthquakes from 1990-1994
become deeper from east to west

The tectonic relationship is also clearly seen in the map of deep earthquakes (figure 2), where under the North Island the earthquakes get deeper to the northwest.  This distribution of all depth bands is parallel to the Hikurangi Trough along which the Pacific Plate dives.  Deep earthquakes also occur in Fiordland and to the southwest of New Zealand, where the Pacific and Australian plates swap their diving and overriding behaviours.  in this area, the Australian Plate does the diving and the Pacific Plate, topped by the south and east of the South Island, is the overriding mass.


Figure 3 - Earthquakes beneath
Fiordland show that the subduction
zone dips towards the east

This change of diving and overriding roles is graphically illustrated by the locations of earthquakes under Fiordland (figure 3).  Here the earthquakes become deeper towards the east.  It can also be seen that the diving angle of the Australian Plate is much steeper than that of the Pacific.

Between the two zones of deep earthquakes, there is a gap in deep earthquake activity which coincides with the Alpine Fault.  This marks the boudary between the Pacific and Australian plates where the two plates push past each other sideways, rather than one plate diving under the other.  They are also pushing together with such force that the Southern Alps are squeezed up along the seam.  The sideways movement means that parts of Nelson and west Otago that were adjacent to one another five million years ago are now 450 km apart, and that in another 10 million years, Haast on the Australian Plate and Christchurch on the Pacific Plate will be close neighbours.

The lack of deep earthquakes along the Alpine Fault zone is because neither plate is diving below the other.  The rocks on both sides of the boudary are too bouyant, and when there is no cold, rigid plate edge descending to great depths, there are no earthquakes generated along the interface.


Figure 4 - New Zealand's shallow earthquakes include those
triggered by crustal deformation and tectonic stresses along
the Alpine Fault.

The Alpine Fault zone is also relatively free from earthquakes between 10 and 40 km deep.  This is due to the pressures from the ongoing sideways grind between the plates creating so much heat that the rocks deform by flowing out of shape, rather than fracture by snapping out of shape through earthquakes.  The distribution of shallow earthquakes, however, shows a somewhat diffuse pattern (figure 4), so clearly the strain can be released by rocks in the top 10+ km snapping in a brittle fashion.

The map of shallow earthquakes also shows that the colliding plates cause the main earthquake type in New Zealand.  A zone of shallow earthquakes along the east coast of the North Island and in Fiordland reflects the subduction boundaries.

Other minor zones lie between Buller and Taranaki, offshore from Wanganui, and in the volcanic zone from Ruapehu north into the Bay of Plenty.  there have been earthquakes in Northland (1963 particularly) and eastern Otago (1974), but these places are much safer from earthquakes than the rest of the country.


Figure 5 - Earthquakes shallower than 40 km between January 1990 and February 1993

New Zealand has about 200 earthquakes a year that are big enough to be felt by people nearby, but these are mostly quire small (figure 5).  Most earthquakes that are big enough to damage buildings are of magnitude 6+, and we generally only have one of those a year.  A magnitude 7 quake accurs about once every 10 years, and a magnitude 8 once a century.

Unfortunately, earthquakes do not occur evenly through time:  they may happen in clusters.  Return-periods are averaged out over many years (with the help of paleoseismologists and scientists in countries with centuries of earthquake records), so they are no cause for complacency.  The averaging procedure that declares a magnitude 6+ 'about' every year really only means that within a hundred years, there is likely to be about a hundred of them.


Looking south along the Alpine Fault, New Zealand's onshore plate boundary.
Milford Sound is in the far distance.