Canterbury's unique geological setting has led scientists to believe the earthquake sequence it has experienced could likely not happen anywhere else in the world.
New research published in Nature Geoscience this week challenges the common assumption that the strength of the Earth's crust is constant by demonstrating that energetic quakes, such as those in Canterbury, can cause widespread weakening of the crust.
GNS scientists had previously considered the heavy ground acceleration recorded at the time of the 7.1 Darfield Earthquake in September 2010 - the result of strike-slip faulting within the Pacific Plate, near the eastern foothills of the Southern Alps - as extremely rare.
The force of the February 22, 2011 quake, which killed 185 people but was considered an aftershock, was so great it was considered statistically unlikely to happen more than once in a millennium, and far exceeded the loading extremes that New Zealand buildings were designed for.
Our codes require buildings to have a 50-year design loads to withstand the loads of a 500-year event, but early reports indicated that the ground motion that afternoon was beyond even 2500-year designs.
Scientists, led by seismologist Martin Reyners of GNS Science, had initially set out to determine the three-dimensional structure of the crust under Canterbury by using a technique called seismic tomography - similar to a medical CAT scan.
This helped to get more accurate aftershock locations, to better define the many smaller faults that ruptured.
Instead, they found that rock properties had changed significantly over a wide area around the Greendale Fault, which ruptured on September 4, 2010 producing a magnitude 7.1 quake.
"This finding was entirely unexpected, but it explains why the main shock released so much energy," Dr Reyners said.
Previously, scientists had assumed that the strength of the Earth's crust remains constant during an aftershock sequence.
The Canterbury quakes had their genesis 100 million years ago when very strong rocks became emplaced under Canterbury.
"It is important to realise that the Canterbury earthquake sequence was very unusual, with energetic earthquakes producing some of the strongest vertical ground accelerations ever seen in an earthquake.
"This is a result of the unusual rock structure of the region. There will be few other places in the world where a similar earthquake sequence might occur."
Most of the quakes in the two-year-long Canterbury sequence released abnormally high levels of energy; this was consistent with the ruptures occurring on very strong faults that store energy slowly and gradually and are hard to break.
Dr Reyners said post-quake analysis such as this research was important as it helps to understand how strain builds up in the crust and how it is released during earthquakes.
"But to do that accurately, we need to understand the types of rocks that exist at depth. Strong rocks store and release strain differently to weak rocks."
The research involved analysing the seismic waves produced by 11,500 aftershocks in Canterbury.
This enabled the team to build a 3D picture of rock structure to a depth of about 35km below the surface.
Normally rocks become hot and 'plastic' at depths of about 10km.
However, the researchers found that strong, brittle rocks continued to a depth of about 30km under Canterbury.
This unusually thick and dense slab of rock helps to explain the long and energetic aftershock sequence in Canterbury.
Seismic energy would have dissipated more quickly in weaker rock.
The researchers are continuing their work and are now focussed on determining how widespread this strong rock unit is in the lower half of the South Island.
"This is important for defining the earthquake hazard for people living between mid-Canterbury and Southland."