Experts: South Island seismic risk needs re-evaluation

Scientists  behind the first big compilation of research into the  Christchurch earthquakes say the seismic risk to the whole of the South Island’s east coast needs to reviewed — and one group of American researchers has warned of risks for Dunedin.

Credit: Royal New Zealand Air Force P-3K Orion (CC BY)The research published in a special edition of a science journal, Seismological Research Letters (SRL), also included insights provided by Greek scientists into why some specific types of structures —  the mid-rise reinforced-concrete buildings that included both the CTV and PGC buildings in which most of the victims died — performed badly:

Registered journalists can log into the SMC Resource Library to access the research

“The inter-story drift demands in the CBD were particularly damaging for all types of structures  but especially catastrophic  for mid-rise reinforced concrete buildings on shallow foundations”.

And engineering researchers at Auckland and Canterbury universities, working with Japanese engineers, noted that “the scale of the damage experienced in Christchurch following the 2010 and 2011 events was unprecedented and may be the greatest ever observed in an urban area”.

A Milan Polytechnic  paper noted “We still cannot rule out the possibility that smaller aftershocks triggered the Christchurch earthquake as a secondary aftershock with a larger magnitude”,  while Cornell University researchers revisited  the theory that the September Darfield earthquake  “likely expedited the timing of the Christchurch earthquake”.

And the Cornell researchers warned that the quakes demonstrated “the need for reassessment of seismic hazards in the eastern South Island” The volcanic structure underlying Bank’s Peninsula — from between 6 million and 11 million years ago —  may have buttressed the quakes.

This idea had already been canvassed separately earlier in the month by GNS Science researchers in EOS Transactions, who said: Banks Peninsula may have  acted as a reflector  in the September quake, increasing the amplitude of the energy waves hitting Christchurch.

And in the February quake, the peninsula may have acted as an oscillator, “extending the duration of the shaking in Christchurch” . The Cornell researchers go on to warn: “Unmapped, potentially seismogenic faults may exist in association with other volcanic structures through out the eastern South Island, such as near Dunedin”.

The 19 articles in Seismological Research Letters ran under a preface that said though both the Darfield and Christch earthquakes ruptured previously-unmapped faults, the Christchurch earthquake was especially meaningful because it followed a larger quake that produced less damage and no deaths. (Some of the papers are available to registered journalists in the SMC Resouce Library, for otehr contact the SMC).

“This earthquake was remarkable on several counts,”  says the guest editor of the SRL edition, Erol Kalkan, a research structural engineer and manager of the National Strong Motion Network with the US Geological Survey (USGS).  “The ground motion was much larger than previously recorded, the high intensity of shaking was greater than expected, particularly for a moderate size earthquake, and the liquefaction-induced damage was extensive and severe within the central business district (CBD),” said Dr Kalkan.

Much of Christchurch was formerly swamp, or was built on fine silt or dune sand in the  estuaries and lagoons that were drained as the area was settled. In some areas of the city built on loose silt, sand and gravel, widespread liquefaction-induced damage within the CBD has required 1000 buildings to be demolished — as laid out in one of the papers, led by Canterbury University researcher Misko Cubrinovski.  Each building damaged  provided “critical insights” into the performance  of structures and foundations on ground that could potentially liquefy.

The  editor-in-chief of SRL, Professor Jonathan Lees — who teaches geosciences at the University of North Carolina —  said that compared to the earthquake that destroyed much of Haiti, the scale of disaster in New Zealand might seem small:  “Christchurch, however, was constructed using much better technology and engineering practises, raising a very sobering alarm to other major, high density western urban centres.”

The 19 papers fell into three main groups:

Liquefaction, shaking and structural damage
Eight papers focus on the structural and geotechnical damages associated with the strong ground motion shaking, comparing differing levels of soil liquefaction and the corresponding structural performance of buildings, lifeline structures and engineering systems. The authors provide a detailed catalogue of damage to stopbanks, bridges and multi-story buildings, including stark contrasts in damage due to differing levels of liquefaction.

Building for the future:
Three papers concentrate on recorded strong ground motions and their engineering implications, and the earthquake and its aftershocks will have long-lasting, significant impact on the understanding of amplified ground motion, and on engineering practices — leading to profound changes in New Zealand’s building code, said Dr Kalkan.

“Many urban areas are built over soft sediments and in valleys or over basins, for example the San Francisco Bay Area and Los Angeles Metropolitan,” Dr Kalkan said. “These are urban areas that sit atop geological features that may exaggerate or amplify ground motion, just as Christchurch experienced. The question is how to apply or account for such significant, higher-than-expected ground motions, as seen in Christchurch, when evaluating the design of existing and new structures.”

Seismological studies:
Comparisons of different fault and interface models, and ways of interpreting the data, including the  potential contribution of the stress change during the Darfield event to the eventual
rupture of the Christchurch quake. New Zealand researchers also evaluate how the complex earthquake sequence of the region likely has arisen through reactivation of a mixture of comparatively newly formed and older inherited fault structures.

The Science Media Centre gathered comment from earthquake experts on the compiled research.

Dr Erol Kalkan, a research structural engineer and manager of the US Geological Survey’s National Strong Motion Network, who edited the special edition,  told the SMC:

“From a purely scientific perspective, earthquake is good news.  Today we have so many sensors installed picking up various signals from seismic events, that analysis of the data boosts our knowledge … in that sense, both Darfield and Christchurch quakes have provided both challenges and opportunities for the engineering and earth science communities.  Scientists … were able to enhance our understanding of the behaviour of Greendale fault responsible for nucleating these two shallow events.  This fault was previously unknown, and it was a complete surprise for us.  Technically, Christchurch and Lyttleton could have been better prepared if we had prior knowledge about location of this fault and it’s potential to create large earthquakes”.

“The most intense shaking experienced during earthquakes generally occurs near the rupturing fault, and decreases with distance away from the fault.   Scientists (in this special issue) have studied the local geologic conditions known as “site effects” as one of the contributors of this difference in shaking intensity in Canterbury region.  Combining this information with estimates of where and how often earthquakes will occur would allow for better estimates of how intense shaking will be during future earthquakes”.

“This special issue not only fuels the scientific knowledge production addressing Greendale fault, the observations and scientific results on significant level of ground shaking reported are extremely important input to create both scenario and probabilistic seismic-hazard maps.  The former show the distribution of ground motions for a possible future earthquake; the latter reveal peak ground motions a site is likely to experience during a specified period in the future.  Data gleaned from these maps is directly incorporated into building codes, prepping future structures for the next great quake and possibly saving lives and dollars when the ground finally decides to shake things up”.

“One of the highlights of both Darfield and Christchurh events are the observed widespread liquefaction and lateral spreading due to soft sediments of Canterbury plains and high ground water level, which were responsible for many damaged buildings and bridges due to instability of their foundation. To prepare for future earthquakes and determining their potential aftermath, we need to know how intense the shaking can be and how often we are likely to experience it.  Now that the researchers will have more information about the behaviour of Greendale fault and Canterbury region soft-sediments and its deep basin structure, which result in amplified ground shaking, as well as liquefiable areas that they can begin to take their science to the next level”.

Geoscientist Professor Kevin Furlong, of Penn State University, who was working in Christchurch when the September  quake hit,  said:

“I found the suite of papers from Misko Cubrinovski’s group to be really interesting and important. Certainly the liquefaction story is probably the really the big one, and the one of most immediate importance to New Zealanders. These papers have identified the character of liquefaction, patterns and amounts, and very importantly to my mind identified the liquefaction and other ground failure behaviour in the CBD. That was something I hadn’t seen before – the focus had been more in the residential regions. From my perspective the results from these various papers are really important considerations as rebuilding decision are made”.

“In terms of the seismological studies …  I find them interesting, but there is little new in them that hasn’t already been in the ‘news’ . Also these sorts of studies benefit from having competing models assessed, so once (and if) the data sets are made available to the wider community we will have a better sense of how robust the various conclusions are. Much of this will be arguing about details, but there are different approaches to analysing and modelling these sorts of data, and having various groups test these and other models will be beneficial to understanding the long term implications of these events for earthquake behaviour in general. There have already been some cross testing of models and assumptions. But much of the papers is again ‘old news’ and didn’t really change our views of what transpired”.

“The suite of ‘strong motion’ papers are very interesting but mostly pretty technical and probably of more interest to the engineering seismology community. I think it is safe to say that the data sets collected are going to be heavily used by that community to test sensitivity to various geological settings (basin shapes/sediment thickness; mechanical properties of the soils, etc.)”.

“I do think that the real potential from these studies (and the data collected) is helping us better calibrate our thinking about the extent of influence from individual events. Having the two earthquakes observed by common instruments is very important scientifically. But this does again reinforce the point that it is critical that a variety of research groups have access to the same data sets. Different analysis approaches are better suited to different questions and these events will have their lasting impact in how the knowledge gained here can be applied elsewhere – and that requires validation”.

In terms of the role that the Banks Peninsula may have played in the patterns and amplitudes of string shaking in Christchurch, I think that the concept is a reasonable hypothesis, however the specifics of how it affects things is yet to be determined. Detailed 3-D modelling of that effect has not been undertaken, so although it is reasonable, how to use that concept in any form of planning (vis-a-vis Dunedin) is still highly speculative. It is also important to remember that any role such a structure may play is dependent not only on its existence but also the spatial relationship between the earthquake generating the energy and the structure.

Although there are clearly some sites that had extreme ground shaking, the overall patterns of ground shaking … are similar to what would be expected from various models used internationally. The key point here is that the data set collected in Christchurch is very high quality and will help to refine these global models (but not dramatically change them).

Prof  Martha Savage, Professor of Geophysics, Victoria University of Wellington, who was not involved in the research published by SRL, said:

The vast number of papers on a wide variety of topics including geodetic and seismological determinations of fault rupture characteristics, analylsis of earthquake locations and liquefaction, questions as to whether and how the earthquakes triggered each other, and performance of buildings and bridges in response to ground motion and liquefaction. It shows the wide level of international, scientific interest in this earthquake sequence, which began with the Darfield earthquake more than one year ago. Authors come from at least six different countries, with just over half the articles including New Zealand authors”.
“The major questions addressed either directly or indirectly by almost all articles are those that may also be uppermost in the general public’s minds: Why did such a small earthquake cause so much damage, and what can we do to prevent future damage? Different articles focus on different aspects, but a summary is that the major causes of damage were: 1) the proximity of the earthquake hypocenter (starting location) to the central business district; 2) the rupture moved rapidly toward the center of the city, piling up energy in front of it much like an airplane that travels over the speed of sound sends a shock wave outward in front of it; 3) for the overall size of the earthquake, it released an unusually large amount of energy in the frequency range that can most easily destroy buildings; 4) the soil in much of the Christchurch area is soft and full of water, leading to conditions almost perfectly suited to liquefaction”.

“Many studies tried to determine the exact pattern of the rupture of the earthquake, with slices in the earth showing which parts had the highest slip. This is not a simple process, and many different types of data are used, including measurements from seismometers continually measuring the rapid motion of the ground; surface measurements of displaced (broken) features such as hedges and road, satellite and airplane photos used to determine the same types of displacement, and GPS measurements from receivers that are either continuously monitoring the ground’s movement or were temporarily installed but re-occupied over various time periods. There are also many different techniques used to use these measurements at the surface to try to determine the measurement beneath the ground (so-called “inversion”).  Naturally there are differences in the results, but reassuringly they all have some basic similarities: the Christchurch earthquake occurred on a thrust fault and ruptured towards the city, with a maximum motion of about 2 metres, located at a fairly shallow depth of 3-5 km beneath the surface. Several articles also discussed whether the stresses from the Darfield earthquake caused the Christchurch earthquake to be brought forward in time. They all concurred that this in fact did happen.

“Another item of interest for the seismological community was the paper by Fry, Benites and Kaiser detailing a “slapdown phase”, a recently-discovered phenomenon occurring only when very large motions are experienced in areas with soft soil. Here the ground acts almost like a trampoline, throwing itself upwards at accelerations higher than that of gravity and falling down with an acceleration equal to that of gravity. This phenomenon makes seismic records look “lopsided”, with the upward acceleration up to several times higher than the downward acceleration, in contrast to most seismic records that overall show generally equal motion in both directions”.

“Personally, an article I found particularly interesting was by a group from Cornell University in the US . They used two types of satellite measurements to determine the fault ruptures of all the major earthquakes. Particularly their use of optical satellite images was relatively new: with high resolution satellite images and image processing software they were able to map the ground movements from the Darfield earthquakes to a degree that was surprisingly similar to the movements measured painstakingly by geologists on the ground . Barnhart et al. also suggested that the 22 February and 6 June aftershocks were both on nearly identical fault planes. Finally, their inversions of the Darfield and 22 February earthquakes most strongly suggested that the stresses from the Darfield earthquake caused increases of stress in the area of the 22 February earthquake, making it more likely to occur”.

“The take-home message of this set of articles is that the Canterbury earthquake sequence brought out a wealth of data and the use of this data is allowing the international seismological community to find out new phenomena under the ground and to answer questions of concern to average citizens. Almost every big earthquake provides surprising new observations that need to be understood and that can help us to prepare better for future earthquakes. Thus, the science of seismology is still young and such research must be continued if New Zealand is to keep its place as one of the top locations in the world to study earth science, and to provide its citizens with the best ability to prepare for future disasters”.

Auckland University researcher  Quincy Ma, who lectures in structural dynamics and structural analysis:

“Christchurch has been and still is an open-air laboratory in understanding the origin, effects and natures of earthquakes… Christchurch has been likened to a massive $30 billion experiment and all that is required now is for engineers and scientists to collect the test data. As insensitive as the statement first appears, this might be some of the good that we can make of the situation to ensure future life losses are minimised. Christchurch is perhaps the best documented devastating near-fault earthquake to hit an urban centre, this definitely has relevance in NZ and overseas. There are still many experiment that can be done with the remaining buildings, I hope they do not too hastily demolish everything. In Taiwan, they have preserved a number of damaged buildings and turned them into museum for education purposes.

“It seems to me from the SRL and previous publications that we now have a reasonable fault model which explains how the September earthquake happened and consequently faulting mechanism for the February event and etc. This enables scientists to predict what might happen for the Christchurch micro-seismo-region in the near future. And from recent reports, the news is good, it all looks to have pretty much settled down. This whole experiences have brought to the public’s attention that while we have a general idea where most fault lines are concentrated, it is not inconceivable for some to be at unexpected locations, and in Christchurch’s  case, at extremely shallow depth, causing devastation. Hopefully, the Royal Commission would provide guidance on how to balance this level of uncertainty, risk and the cost of doing something.

As an engineer, not a fully qualified seismologist, I disagree with the conclusion there is sufficient evidence that the Christchurch earthquakes are in any way related or caused by the presence of the volcanic zone. It is to me a far-fetch to then suggest that similar structure could exist and to which similarly devastating earthquake could occur. There is perhaps just as likely that seismogenic faults exist in any other part of the country”.

“Christchurch is a wake up call for the public to evaluate what is an “acceptable level of risk” for them”.

The papers published in the first major scientific compilation from the  February quake included half a dozen in which staff from state science company GNS Science were either lead authors or contributing authors. Key points from these included:

Sibson, Ghisetti, and Ristau paper :  Quake sequence arose from tectonic stresses on a mix of older inherited and comparatively newly formed fault structures.
Some fault ruptures were re-activations of inherited basement faults, but other comparatively low-displacement fault structures may be newly formed.

Overall, the fault system responsible for the Canterbury quake sequence appears to be controlled by the orientation of the tectonic stress field in the upper crust rather than conforming with local plate boundary kinematics.
On this basis, the quakes can be regarded as intra-plate quakes remote from the main Alpine-Marlborough fault system defining the onshore plate boundary.

Holden paper:

– All evidence points to a high-stress drop event.
– Had rupture velocity of 3.2km/s.
– Quakes in Canterbury are particularly energetic.
– The region is characterised by a very dehydrated and brittle bedrock structure – the Hikurangi Plateau.
– After a large quake this bedrock promotes strain release by a long sequence of aftershocks rather than aseismic slip.
– The Christchurch area also features the remnants on volcanism from the extinct Banks Peninsula volcano.
– The intrusion of the volcano has segmented faults in the Christchurch area and may also have brought the brittle bedrock closer to the surface.
– This helps to explain highly energetic quakes such as the February quake.

Bannister paper:

– Paper is based on analysis of 2177 aftershocks following the February 22 quake.
– Providing higher accuracy in location and depth of each aftershock.
– Most aftershocks after February quake occurred at between 3km and 8km depth and within a few kilometres of the inferred February fault plane, but not clearly on that fault.
– Possible reason for this is has been very little post-February slip on the Port Hills Fault.
– A lack of viscous deformation at depth has probably encouraged high levels of stress buildup in the bedrock under the Canterbury gravels.

Fry, Benites, and Kaiser paper:

– Investigated the reasons for the strong vertical and horizontal ground motions in the February quake

Fry and Gerstenberger paper:
– More work is needed to understand the relative contributions of stress drop and directivity to ground motions. This has implications for the NZ national seismic hazard model.

Further  research to be published:

  • A New Zealand Journal Geology and Geophysics (NZJGG) special issue in 2012 will cover a range of scientific and engineering topics, including the earthquakes’ geological and historic contexts, ground motions, instrumental and field observations, effects on buildings/structures, groundwater responses.
  • There will also be a special issue  of the Bulletin of the New Zealand Society for Earthquake Engineering just before Christmas or in early January dedicated to the 2011 February Christchurch earthquake including seismological characteristics and building performance, liquefaction, landslides and infrastructure.