NHRP / Hazard themes / Geological Hazards / Seismology / Seismology Highlights 2013-14

Seismology Highlights 2013-14

Earthquake Ground Motions: Learning from the Canterbury Earthquakes

Next >> Modelling Earthquake Rate Forecasts

Back to Seismology

Ground motions recorded during the Canterbury earthquake sequence reached some of the highest levels documented worldwide, and the ground shaking varied across the region in terms of its amplitude and frequency. Being able to accurately predict ground motion characteristics for a specific location is necessary to build resilience to earthquake hazards. A number of different factors are thought to explain the observed ground motion characteristics: these include the earthquake source, the wave propagation path and the local ground conditions (‘source, path, and site’). The excellent coverage around the epicentre by the GeoNet seismometer network allows us to investigate these factors for Canterbury and improve and update the current ground motion prediction models. In this joint project of GNS Science and University of Canterbury, we are using several methods to quantify the influences on ground motion.

Working with Adrien Oth from the European Centre for Geodynamics and Seismology, we have been able to introduce state of the art approaches to these studies. Our preliminary results suggests that earthquakes in Canterbury tend to release higher than average energy for their size (i.e. they exhibit higher stress drop; see graphic),

Kaiser stress drop original fig

and this is particularly true for events clustered near the tip of faults that ruptured previously during the sequence. The higher stress drop is thought to be related to the tectonic environment, where rare events, such as the Canterbury earthquake sequence, occur on strong faults within rigid brittle crust. Earthquakes in other historically more active regions of New Zealand might behave differently, and we are testing this in the recent Cook Strait earthquake sequence.

Given that many urban areas of Canterbury are located on deep and/or soft soils, our analysis includes use of models representing the 3D response of liquefaction-susceptible soils which we developed as part of this project. In addition, modifications to ground motion prediction equations have been developed and included in the National Seismic Hazard Model to provide more accurate estimates of shaking for structural design and liquefaction assessments.

In the past, traditional hazard assessment has relied on recorded data from strong-motion sensors or seismometers. Now we are able to create accurate simulated data of ground motions for various Canterbury scenarios. This approach utilises the region-specific ground motion characteristics determined above (‘source, path, site’). When compared against the recorded data for the large events of the Canterbury sequence, the simulations can accurately estimate ground shaking levels and characteristics at new sites within Canterbury. We can retrospectively assess past performance of structures or landmass under different levels of shaking or simulate ground motions of future scenarios. The use of synthetic approaches is cutting edge and paving the way of the future for seismology.

Left: 'Source, path and site' effects influence ground motions which can be measured by strong-motion sensors or seismometers. These effects can vary between different regions or even different nearby locations. Right: We can simulate ground motions for a given earthquake and a given location by drawing on historical data from events like the Canterbury earthquakes.

This research is funded by the 'Lessons Learned from Christchurch," a Canterbury Earthquake Research Strategy programme, and is jointly led by Anna Kaiser (GNS Science) and Brendon Bradley (University of Canterbury) with contributions from colleagues at these institutions and overseas.


Oth, A.; Kaiser, A.E. 2013 Stress release and source scaling of the 2010-2011 Canterbury, New Zealand earthquake sequence from spectral inversion of ground motion data. Pure and Applied Geophysics, doi: 10.1007/s00024-013-0751-1 (16 p.)

Back to