Scientific response to shakes explained

The following commentary by Associate Professor John Townend from Victoria University's School of Geography, Environment and Earth Sciences was originally published in The Press on 28 November. The geophysicist explains the scientific response beyond the recent magnitude 7.8 earthquake to strike New Zealand.

Displacement of a farm track
A farm track along the Kekerengu Fault displaced horizontally by about 10 metres, and vertically by about 2 metres. Will Ries (GNS Science) is using a GPS unit to measure the offset of the displaced track. Photograph: Tim Little.

The challenge for geoscientists in the aftermath of a large earthquake is determining where aftershocks are likely to occur, how often, and what their effects will be. To do this, scientists need to know what has happened, what is happening now, and what past experience of aftershock sequences tells us is likely to happen in coming days, weeks, or months.

Working out what is happening is a team effort—many different types of observation and scientific expertise is required. Every second, data from more than 450 seismometers and 200 global positioning system (GPS) instruments spanning the country pour into GeoNet’s computers.

Separately, the levels of ground-shaking recorded at sites throughout New Zealand are also recorded, forming the basis for engineering assessments of buildings and other infrastructure in the event of strong shaking. Field parties map the location and amount of fault slip that has occurred, and the locations and scales of landslides.

Determining what has happened involves a combination of painstaking fieldwork in often arduous conditions, and extensive computer analysis of aerial and satellite imagery.

Geodesists begin analysing data from GPS recorders, which show how the ground has moved, in three dimensions, and in conjunction with seismic data, field observations and satellite imagery delineate the faults that have slipped. Before-and-after images of the Marlborough region have already yielded astonishing pictures of ground deformation and show that the Kaikoura earthquake involved slip on multiple separate faults.

Models of where and how much slip has occurred during a large earthquake are used to investigate how adjacent faults have been loaded or unloaded. Just as one person getting off a crowded bus will cause other passengers to shuffle about and find more comfortable positions, so slip on a fault causes stress to be distributed in the surrounding area. This stress redistribution process increases the stress on some faults, bringing them closer to the point of failure, and reduces stress on others. The size of the Kaikoura earthquake causes stress changes over a large region—much of the northern South Island and Cook Strait—and an immediate priority for scientists is working out how other large faults have been affected.

It’s also important to understand the hazard posed by ongoing aftershocks. Seismologists are often quoted as saying that the number of earthquakes of a particular size in a region decreases by a factor of ten for every one magnitude unit increase in size. In other words, for every ten magnitude 4 earthquakes, scientists expect to record approximately one magnitude 5. This rule of thumb, known as the Gutenberg-Richter relationship, holds well for earthquakes in most parts of the world, and enables an estimate to be made quite quickly of how many aftershocks of a particular size are anticipated following a large earthquake. The timing of those aftershocks is reasonably well described by the Omori relationship, which states that the average frequency of aftershocks decreases systematically as time goes by.  

Following a large earthquake, GeoNet scientists compute aftershock forecasts regularly, revising each forecast on the basis of what earthquakes have been recorded so far.

Knowing what earthquakes to expect is only part of the challenge. What is of more importance to engineers and civil defence authorities is an assessment of the amount of shaking different parts of the country may experience. To calculate this, seismologists combine the aftershock probabilities with models describing the decay in shaking with distance from an earthquake and the effects of different soil types and geographic features.

GeoNet uses a combination of probabilistic earthquake forecasts and quantitative assessment of specific scenarios to communicate the evolving hazard posed by aftershocks and other possibly large earthquakes. It has to be remembered that these assessments are made objectively on the basis of a range of instrumental observations and past experience of New Zealand earthquakes. However, complexities of the earthquake source, the ground deformation, and the ensuing patterns of aftershocks make it impossible to know exactly what will eventuate.

In the aftermath of a large earthquake, it is natural to be concerned about what may follow. The analysis underway by dedicated New Zealand scientists working closely with colleagues overseas provides a strong basis for planning and preparations—as individuals, as families, and as a society—and is something all New Zealanders can take comfort and pride in.