Detailed analysis of last September’s Darfield earthquake has confirmed that it was a complex event, with many similarities to the Haiti earthquake in January last year.
Friday, 15th April 2011
The magnitude 7.1 earthquake centred near Darfield on the morning of the 4th of September 2010 has intrigued seismologists since the moment earthquake information began flowing in to GeoNet’s data centres. While an Alpine Fault earthquake was expected, a major quake under the Canterbury Plains came as something of a surprise.
Initial analysis of information supplied by New Zealand seismometers showed the earthquake was a reverse faulting mechanism, with one side of the fault pushing up the fault surface and over the other side. However, data from overseas seismometers located thousands of kilometres away suggested a strike-slip event, with the land on one side of the fault moving sideways with respect to the other. The reality was a bit of both.
It is now thought that some major earthquakes are not simple single-rupture events but that they consist of several seismological events that occur in rapid succession (that is within seconds, not minutes). The events are so quick that human observation cannot distinguish them apart because they occur during the period of heavy shaking that might continue for a minute or more thereby masking the individual elements of the event. The magnitude 7.1 Darfield Earthquake was a case in point, consisting of several related seismic events occurring in a short space of time. Most people thought they had experienced a single large earthquake whilst a smaller number were able to distinguish two of the more distinct events, depending on where they were located.
Instrumentation and rapid data collection allows these earthquake events to be examined second by second, with computers being used to filter out the effects of the first event to see what subsequent activity had occurred.
In the case of the Darfield earthquake, GeoNet’s seismic network had comprehensive coverage of the region, and Global Positioning System (GPS) monuments were able to provide further information about changes in the landscape before and after the quake. Serendipity played a large part too.
The anticipated Alpine Fault earthquake is expected to be a large event and could consist of three or four magnitude 7 earthquakes over a period of minutes adding up to a combined magnitude of 8 or more as the fault ruptures in successive sections triggered by the previous event. Whilst the Alpine Fault may rupture like a zipper opening northward or southward along the fault, it is expected to occur over a slghtly longer time frame than the Darfield event.
In order to study the Alpine Fault quake, the University of Canterbury had deployed a network of strong motion detectors on the Canterbury Plains. As luck would have it, some of these instruments were in amongst the activity caused by the Darfield earthquake, providing valuable additional information on what occurred during the seismic event.
Analysis now shows that the Darfield event began with the rupture of a blind fault to the north of the Greendale fault at 4:35 a.m. on September 4th. It was a reverse faulting earthquake of about magnitude 5.3 with one side of the fault pushing up over the other but it did not break the surface, and hence is considered a “blind” fault.
Ten to fifteen seconds later the movement triggered the previously unknown Greendale Fault which did rupture to the surface over a length of about 30 kilometres. This was a strike-slip rupture, with opposite sides of the fault moving horizontally to the right by about four to five metres with respect to each other.
The movement along the Greendale Fault triggered another blind reverse fault at its western end, and possibly a blind reverse fault at its eastern end.
The Haiti earthquake of January 2010 was similar to the Darfield event in many of these respects. Both quakes were of similar magnitude and depth, and the Haiti quake consisted of a number of nearby faults rupturing very quickly one after the other. But it has to be pointed out that the Haiti event had some dramatic differences too. Maximum shaking intensity in Haiti reached MM9 and affected nearly 600,000 people, with a further 2 million people exposed to shaking of MM8. In Canterbury, perhaps 50,000 people were exposed to MM9 shaking between Hororata and Lincoln, with a further 400,000 in Christchurch urban areas subjected to MM8 shaking.
The complex activity at Darfield did not end there, however. The 3,800 aftershocks analysed by GeoNet show that most of the aftershock activity has occurred along the Greendale Fault. But there are also three distinct clusters of aftershocks – one at each end of the Greendale Fault and one slightly north of the Greendale Fault, on the eastern side of Darfield.
Speculation over whether the Darfield quake was a multi-quake event began the same day that the earthquake struck. Dr Kevin Furlong, on sabbatical from Pennsylvania State University, told NZPA that there had been a magnitude 5.8 foreshock about 5 seconds before the main earthquake, and that the main shock had two pulses of energy in it. He thought that the foreshock had awakened a lot of sleepers just in time to experience the main earthquake. GeoNet’s earthquake catalogue contains a record for the magnitude 7.1 earthquake, but no data for the closely related events mentioned by Dr Furlong or GeoNet News.
For those interested in the scientific detail, there has been an interesting discussion on the all-geo website about the Darfield earthquake and its ramifications for nearby faults. It is, of course, a discussion and wanders across a number of points, so don’t expect to read a nicely laid-out article, and make an effort to read the comments from contributors in Canterbury and the greater scientific community.
One intriguing matter discussed relates to unusual effects seen in the Lyttelton Road Tunnel at the time of the earthquake. Just before the Darfield quake struck, a shockwave was seen to pass through the 2km-long tunnel in an east to west direction – that is in a general direction toward the earthquake epicentre. The shockwave looked similar to a very fine dust cloud but dissipated too fast to be dust.
Cctv camera footage was examined after the earthquake and it was found that shaking started at Lyttelton at the eastern end of the tunnel before it started at Heathcote, at the western end. Bearing in mind that the epicentre was north-west of the tunnel, logic would suggest that shaking should have been recorded at Heathcote before being felt at Lyttelton as the s-wave from the earthquake propagated outward.
Another fascinating effect was observed. The cctv footage showed some kind of lightning effect coming out of the ground, and the poster (who claimed to be in the control room for the tunnel at the time the quake struck) mentioned that the police had reported seeing the same phenomena.
The reporting of lightning-like phenomena in conjunction with a large earthquake is not new, but it is not common. Weird lighting effects in association with earthquakes in New Zealand were reported as early as the 19th century. Explanations range from methane gas escaping and igniting, to electrical discharges from quartz being suddenly placed under extreme pressure. Nevertheless, it would be interesting to see the tunnel cctv footage, and have the phenomenon followed-up.
What is unexpected and unusual, is the cctv footage seeming to show a shockwave passing through an underground structure toward an earthquake epicentre before the earthquake struck. The video captured could be unique. Timing is, of course, everything. It would have taken several seconds for the s-wave to reach the Lyttelton Tunnel from Darfield, so analysis may actually show that the shockwave occurred at the same time or just after the triggering of the foreshock, or the main shock, or, as it appears, just before one or both.
Unfortunately, discussion about the tunnel observations has gone quiet, and the Darfield event has been overtaken somewhat by the shocking earthquake of February 22nd. No doubt the matter will be re-examined as time permits.
Research into the Canterbury earthquakes is continuing with the sequence that began with the Darfield magnitude 7.1 quake being studied in conjunction with the magnitude 6.3 event that was triggered near Lyttelton in February. Some early speculation suggests that the entire sequence has been a triple event with one magnitude 7 and two magnitude 6 earthquakes. The first of the magnitude 6 events may have occurred to the west during the Darfield quake with the second event being the disastrous Christchurch quake five months later.
[ Sources: GeoNet News Issue 14 March 2011, all-geo.org website, eyewitness reports, news items. ]