View this blog in blog mode
What is a “ghost quake”?
“Ghost” quakes appear on our network typically after a large regional source earthquake. We have very sensitive seismic equipment that picks up the various waves that earthquakes create and we can pick up these waves even if it is very far away. For example, we also picked up seismic waves from the Alaska quake this morning. Our equipment gets confused by these waves and interprets these as being a smaller, locally-sourced earthquakes close by.
Why do we get “Ghost Quakes”?
These quakes are an unfortunate side effect of getting information out to the public as quickly as possible, instead of waiting up to a quarter of an hour for a person to locate and ensure these are authentic earthquakes. This started when we introduced “GeoNet Rapid”; the up side is that the system is highly efficient and quick with earthquake reporting times of just minutes after the quakes.
It is sharply arriving S-waves - that our automated system confuses for P-waves - that are causing the incorrectly reported earthquakes. Our seismologists are easily able to see the confusion and gradually mop up the false earthquakes on the website.
How can we stop “Ghost Quakes”?
We are working on it but it's tricky. If we make the system too picky on the quakes it reports, we might not get rapid information about real earthquakes that occur. If we let it report on whatever information it picks up, we get “ghost quakes”. At GeoNet, we have erred on the side of speedy reporting of everything because we know how important it is to get information as quickly as possible to everyone.
Having said that, we are working on finding out what we can do to better identify them and so exorcise “ghost quakes” from our system. Until then, please be patient with us, and keep up the sense of humour, New Zealand!
(Image: provided by Stuff.co.nz)
It's a question we get a lot here at GeoNet given that all of New Zealand's significant earthquakes in the past few years - Canterbury, Cook Strait and Eketahuna - have struck on previously unknown faults. And the answer is... complicated.
There are currently four simple ways to tell if a fault exists: we can see it (land deformation), we've heard about it (it's in our written or oral histories), we've dug into the ground and can see it in the soils and rocks, or we've recorded an earthquake on it (read below why this isn't necessarily useful for finding large faults). If it's not one of those four ways, it becomes difficult to know if a fault exists in an area and if it can cause a significant earthquake. There is a fifth way but I'll get to that in a moment...
How often do earthquakes happen on previously unknown faults?
Unfortunately the answer is a lot; a magnitude 6-7 earthquake occurring on an already identified fault is the exception rather than the rule. Luckily, the odds of larger earthquakes occurring on identified faults are much better; once we look into magnitude 7.8+ earthquakes, researchers are confident that all faults in New Zealand capable of these sized quakes have already been identified and mapped. Why? Because we can see them clearly in our landscape. A big earthquake requires a fault of significant length; large faults with the potential to generate magnitude 7.8+ earthquakes - like the Alpine Fault - will form a significant feature of New Zealand's landscape and are therefore much easier to find.
Earthquakes are happening all the time - don't they show us where faults are?
As we all well know after the Canterbury quakes, parts of the country can be relatively quiet for decades but still capable of large quakes; we'd need 20,000 years' worth of earthquake records to have the complete picture!
Every earthquake happens on a fault, and with a searchable database of over 480,000 earthquakes we know where a lot of faults are. The trouble is that all these earthquakes don't nicely line up and illuminate individual sub-surface faults. The accompanying map showing just three years of shallow earthquakes in the middle of the country illustrates this. Also, as large quakes are infrequent, only 30,000 of the 480,000 are magnitude 4+ quakes so it's impossible to know whether the small quakes are occurring on a small fault, or they're small quakes occurring on a small portion of a larger fault.
Our current New Zealand active faults database contains 532 mapped faults that are capable of magnitude 6+ earthquakes. Although the amount of known active faults has increased significantly in the last few decades, there are still a large number of unknown faults capable of producing large earthquakes. Some research conducted by Dr. Andy Nicol at GNS Science suggests that there could be as many as 3500-4000 faults around the country capable of producing magnitude 6+ earthquakes, though the vast majority of these faults will be in remote areas of the country.
How come we, the people paid to know where faults are and what will happen, don't know?
The reason that so many faults are still undetected is that faults capable of magnitude 6-7 earthquakes aren't present at the Earth's surface and are therefore difficult to map. Earthquakes generally originate at depths greater than 5km, so only the very large, long-lived faults fracture rock all the way up to the surface. Faults that do not reach the surface are called 'blind' faults, and we aren't the only country with this problem. Other seismically active places around the world also have to contend with significant earthquakes occurring on these unknown blind faults. One famous example was the magnitude 6.7 Northridge Earthquake that struck suburban northern Los Angeles in the early morning of 17th January 1994. The quake resulted in 60 deaths, and more than 40,000 damaged buildings. Although it was widely known that California has many faults, this particular fault and its proximity to the densely populated city was unknown.
Finding all of the potentially thousands of blind faults in New Zealand is currently impossible. Geophysical surveys - similar to rudimentary MRI scans - are the way we traditionally find blind faults. These surveys look to find layers of rock offset below the surface. Although the technology does exists for us to go out and find most blind faults capable of large earthquakes, currently it would be prohibitively expensive to do this over the whole country. Even using this technology, faults which move very infrequently, like the Greendale fault - one of the faults responsible for kick-starting the Canterbury earthquake sequence - are particularly hard to identify as the subsurface rock layers are minimally offset.
Scientists do sometimes carry out such geophysical surveys to understand earthquake risk and to gain a better understanding of New Zealand's subsurface, and generally focus on finding faults that have the highest potential for devastation, especially in areas near high population and/or high likelihood of earthquakes. Geophysical surveys are also carried out by industries (such as oil and gas) and are utilised by scientists looking for earthquake generating faults. Surveys from the oil industry were used to locate faults around the Canterbury Plains (the Greendale Fault was not found, but other faults were known pre-2010). So this approach certainly isn't perfect.
So what do we do all day if we can't tell you where all faults are?
Researchers tackle the problem of unknown faults in a few ways. Firstly, they understand that knowing where these individual faults are only goes so far; we'd also need to know how often large quakes reccur on that particular fault, and when the last big one was on that fault. Research focusing on tectonic regions as a whole can offer more relevant insights into different areas' earthquake potential. We know that areas like Wellington and the Alpine Fault accumulate more strain at a faster rate from the collision of tectonic plates than other areas of the country (and therefore must release this strain more often via earthquakes). Secondly, as large earthquakes are inevitable (but infrequent), researchers focus on ways to mitigate the effects, such as building design and developing resilient communities.
Right now, it is like playing a constant game of catch-up with the earth, knowing that it has billions of years ahead of you and you only have 50 years of research behind you to unravel its secrets. We want to be able to tell everyone one day where all the faults are in New Zealand, when these will rupture and how big they will be, but until we can do this, it is important for all of us to be prepared for earthquakes.
For those of us in the middle of the country, it was hard to miss that mid-2013 to early 2014 was an especially active period when it came to earthquakes. It kicked off in July 2013 with the Cook Strait sequence – which included magnitude 6.0, 6.5, and 6.6 earthquakes – followed by the magnitude 6.2 Eketahuna earthquake in January 2014, and the magnitude 5.2 Waipukurau quake in March 2014.
2013 was also an especially active time for ‘silent earthquakes’, also called slow-slip events, which are similar to earthquakes as they involve fault movement. However, unlike earthquakes which occur in a matter of seconds, slow-slip events happen over weeks to months. In the last decade, slow-slip events have been discovered at plate boundaries around the world and GeoNet’s continuously operating GPS network in New Zealand has enabled the detection of slow-slip events at the North Island's Hikurangi subduction zone.
Last year four slow-slip events occurred around the North Island, including the largest slow-slip event ever recorded in New Zealand - it was offshore from the Kapiti Coast and equivalent to a magnitude 7.1 earthquake. A slow-slip event in Hawke’s Bay, equivalent to a magnitude 7.0 earthquake, was the largest on record for that region. Just like earthquakes, slow-slip events release strain, but without our GPS instruments, slow-slip events would be unreported as they can’t be felt and do not noticeably deform the ground. Nevertheless, the altered stresses around a slow-slip event can impact shallower faults that break via traditional earthquakes.
The extent to which recent slow-slip activity is related to the recent large earthquakes (and vice versa) is an area of active research, as we learn more about these relatively new phenomena (see below).
The slow-slip events of 2013
The Kapiti event started in early 2013 and is still going on. It is New Zealand’s largest slow-slip event ever recorded, equivalent to a magnitude 7.1 earthquake. We have 12 years of records in this region, which show three Kapiti events, each occurring roughly every five years. So far, this current event has as much as 25cm of movement over an area approximately 100km by 200km. The previous event in 2008 was equivalent to a magnitude 7.0 earthquake, but the movement was more concentrated: 30cm of movement over a smaller area near Kapiti Island. The 2013 slow-slip event evolved over the course of the year, moving north-eastward towards the North Island, with movement diminishing in late 2013-early 2014.
In February 2013, Hawke’s Bay experienced the largest slow-slip event for the region in a decade of recording. This slow-slip event occurred over several days and was the equivalent to a magnitude 7.0 earthquake. It was also associated with a swarm of over 100 earthquakes larger than magnitude 3 recorded during the period of slow slip, with the largest being of magnitude 4.8. These earthquakes occurred over a period of hours, and migrated deeper over time.
Two slow-slip events were recorded in the Gisborne region in 2013: one in July, the other in October. Both events were smaller than the Kapiti and Hawke’s Bay slow-slip events. The July slow-slip event occurred to the northeast of Gisborne, whereas the October event occurred to the south. The amount of movement and exact location of these events are not precisely known yet, but a temporary deployment of sensors on the ocean bottom placed offshore from Gisborne are being retrieved this week, and these should have captured how these events deformed the seafloor above. While these sensors are being retrieved, a larger set of ocean bottom pressure sensors and seismometers are being deployed as part of a joint NZ/USA/Japan project. These will offer our best insight yet into Gisborne slow-slip events - read a blog of the deployment here.
The relationship between slow-slip events and earthquakes
There are several examples from both New Zealand and overseas of earthquake swarms accompanying slow-slip events. Because slow-slip events occur over a large area, the amount of stress they transfer to other faults is diffuse. This is unlike a large traditional earthquake that has a large stress transfer concentrated in a relatively small region. For this reason, a magnitude 7.1 slow-slip event is probably not going to have anywhere near the associated triggered earthquake activity that we saw after the magnitude 7.1 Darfield earthquake, but it will increase stress in surrounding areas, and could push an already stressed fault closer to rupture. In essence, it can be the straw that breaks the camel’s back. This is possibly what happened with the January 2014 Eketahuna quake, where the Kapiti slow-slip event loaded stress on the fault that broke. The fault in the Eketahuna earthquake would most likely have ruptured in the near future, but the added stress may have caused it to rupture earlier.
It is important to note a few caveats when looking at the transferred stress events from slow-slip events (or regular earthquakes):
- Slow-slip events do not universally increase stress on surrounding faults, they also relieve stress in some areas, and therefore may postpone an earthquake in an area of decreased stress.
- As we can’t directly measure the plate movement (because it is well below the surface and most often offshore) we have to use models to determine what is going on and how the stress is being distributed. As such, the model will hopefully give a good indication, but will not be perfect, as there are many details that we cannot resolve using solely land-based instruments.
Why we care when slow slip causes no shaking
Before discovering slow-slip events, earthquakes were thought to be the only way the Earth’s crust could relieve the pent-up stresses caused by the moving tectonic plates. With the discovery of slow-slip events, this thinking has been drastically altered, as slow-slip events accommodate a large proportion of the effects of the converging plates without knocking a single ornament off a shelf. Slow-slip events in themselves don't pose any risk to people, but they are a major part of how the tectonic plates move in a subduction zone. The other major part is earthquakes. So if we better understand the slow-slip events, we should better understand the earthquake potential of subduction zones.
The subduction zones where slow-slip events occur (in our case where the Pacific and Australian plates meet) are responsible for generating the world’s largest earthquakes – ‘great earthquakes’ or ‘mega-thrust earthquakes’ – which have a magnitude greater than 8. These types of earthquakes can also produce tsunami with deadly consequences as we’ve seen in recent times in Japan and Sumatra. Scientists think that a future mega-thrust earthquake with a magnitude of 8 or larger is possible on New Zealand’s northern subduction zone – the Hikurangi subduction zone. An earthquake this large would produce damaging shaking for much of the North Island, and could produce a significant tsunami affecting much of the country (and also some coastal regions around the Pacific). However, scientists still don’t know how much, and how often, the Hikurangi subduction zone ruptures in megathrust quakes. The best evidence at the moment suggests they are relatively rare, happening every 1500 years. This is an area of active scientific research in New Zealand, and the more we know about the slow-slip events, the more we understand the subduction zone as a whole, and ultimately the better prepared we, and other nations, can be.
I have recently read an excellent article by John Stenmark that answers this question. It has the added bonus of using examples from our part of the world:
With this week's large earthquake and our joint messages on preparedness, we thought you might be wondering how earthquake scientists prepare their own homes for earthquakes. What's in their emergency supplies? How do they brace their televisions and shelves? How do they strengthen their own homes?
Well, we have asked a few of our earth scientists and here is what we found out ...
GNS Scientists : Earthquake preparations at home
What has happened?
The M6.2 Eketahuna quake struck at 3:52 pm on Monday, 20th January 2014, centered 15 km east of Eketahuna, in the Wairarapa, New Zealand. The quake was felt strongly in both islands, and we have received over 9000 felt reports from the public, with multiple reports of damage from those closest to the quake. The focal mechanism shows it to be a normal fault earthquake.
What will happen next?
In research published in 1994 by GNS Science, a slow-slip event (SSE) was thought to have affected the stress on the faults associated with the Weber 1990 earthquakes. There is currently a SSE beneath the Kapiti coastline, which has been in progress since early 2013. Preliminary calculations of stress change indicate that this ongoing Kapiti SSE may be causing changes in stress beneath the Tararua and Wairarapa region. Research relating to SSEs and their relationship to earthquakes is ongoing here at GNS Science and elsewhere around the world.
The Eketahuna quake struck at 3:52 pm on Monday afternoon, and was centered 15 km east of Eketahuna, under the south-east of the North Island. The quake was felt strongly in both islands, and we have received over 9000 felt reports from the public, with multiple reports of damage. The focal mechanism shows it to be a normal fault earthquake.
An aftershock sequence is ongoing following this earthquake and includes numerous magnitude 4 events. As with any aftershock sequence, we cannot rule out the possibility of occurrence of future larger earthquakes.
What are the probabilities of future significant earthquakes?
Up-to-date aftershock forecasts are now located on the 'Future Scenarios and Aftershocks' page in the News Section of our website.
Current depth and location estimates of this event place it below the interface between the subducting Pacific plate and overriding Australian plate. Maximum peak ground acceleration recorded by GeoNet instruments is about one quarter of the acceleration due to gravity (0.26 g). This measurement was recorded in Woodville. Relatively strong shaking was also recorded on the Kapiti Coast (e.g. up to 0.2 g in Paraparaumu) whereas accelerations recorded in Wellington city were less than 0.05 g. In comparison, ground motions recorded in Wellington during the recent Cook Strait earthquakes ranged up to 0.26g.
The seismology team at GNS Science is also attempting to understand the earthquakes in the context of recent and ongoing seismicity. Specifically, efforts are aimed at comparing the current earthquakes with a sequence of events that occurred to the north of Eketahuna in the early 1990s. We are also considering possible links to the ongoing Kapiti slow-slip event.
The likelihood of a larger triggered event in the coming weeks is minor. However, it is possible and we should all take this opportunity to review our earthquake emergency plans.
What are the numbers so far?
Numbers of Eketahuna region earthquakes
|Magnitude range||15:52 pm, 20 January 2014 to|
08:40 am, 27 January 2014*
|6.0 - 6.9||1|
|5.0 - 5.9||0|
|4.0 - 4.9||8|
|3.0 - 3.9||75|
|*This table will no longer be updated. Up-to-date aftershock tables are in 'Our News' section|
- GNS Science and GeoNet have deployed three strong motion sensors to the region to supplement the permanent GeoNet network, you can read about the deployment via the GeoNet blog
- Our Landslide response team are currently asking for public reports on landslides,on private property in the areas affected, please send any information to firstname.lastname@example.org
On Friday, 19 July, a magnitude 5.7 earthquake located in Cook Strait was felt strongly throughout the lower North Island and upper South Island. This quake was 25 km east of Seddon and at a depth of 16 km, and was felt from Auckland to Otago, with reports of minor damage such as items falling from shelves reported in Marlborough, Nelson and Wellington. This was the first of a complex sequence of earthquakes that continued to shake the area for the following several weeks, and for which smaller aftershocks still continue. It was followed on the morning of Sunday, 21 July, by a similarly located event of magnitude 5.8, and then 10 hours later, at 5:09 pm, by another of magnitude 6.5.
This quake, known as the Seddon earthquake, resulted in varying degrees of damage to thirty-five buildings within the Wellington CBD, with glass from broken windows falling onto the main thoroughfare. Instances of moderate to serious building damage were also reported throughout the Wellington region, and in the Seddon area. Four people received minor injuries, and it was reported felt from Northland to Otago. Four aftershocks of magnitude 5.0 to 5.5 and about 70 of magnitude 4.0 to 4.9 followed in the period to 15 August.
At 2:31 pm on 16 August, a magnitude 6.6 earthquake centred at Lake Grassmere, 10 km south-east of Seddon, occurred. This was at a depth of 8 km, and caused significant damage in the epicentral region, including land damage and landslips that blocked several roads, and serious damage to buildings. In Wellington, moderate damage was caused to some buildings and some had to be evacuated. Suburban train services were cancelled and there were traffic jams on major roads. It was reported felt from Auckland to Invercargill. Four minor injuries and one serious medical condition were reported. In the period to 5 September, there were 16 aftershocks of magnitude 5.0 or greater, with one of these of magnitude 6.0 occurring on 16 August. The last aftershock to have been strongly felt was a magnitude 4.7 event on 6 December.
Other significant earthquakes were:
- On 1 January, a shallow magnitude 5.0 centred 30 km north-west of Opunake, was felt from Auckland to Nelson, most strongly in the Taranaki region. Minor damage was reported in Stratford. A magnitude 4.3 aftershock shortly later was also felt strongly.
- On 16 February, a magnitude 6.3 centred 200 km north of Te Araroa, at a depth of 290 km, was felt from Bay of Plenty to Otago, most strongly in the north-east of the North Island.
- On 23 February, a magnitude 4.8 centred 40 km south-east of Wairoa at a depth of 70 km was felt in the Gisborne – Hawkes Bay region.
- On 17 March, a magnitude 3.9 earthquake centred 15 km north-east of Auckland, near Motutapu Island, and at a depth of 6 km was felt strongly throughout the Auckland region. There were several reports of minor damage such as items falling from shelves. Close to 14,000 felt reports were submitted for this event. It was preceded a few minutes earlier by a similarly located magnitude 3.1 that was also widely felt in Auckland.
- On 4 July, a magnitude 5.3 at a depth of 5 km and centred 25 km north-west of Milford Sound was felt from Southland to the West Coast, most strongly in Queenstown and Oamaru.
- On 8 July, a shallow magnitude 4.9 centred 25 km north-east of Waipukurau was felt from East Cape to Christchurch.
- On 29 July, a 9 km deep magnitude 4.7 centred 20 km north-west of Culverdon in North Canterbury was felt from Taranaki to Otago.
- On 17 December, a magnitude 6.2 centred 30 km off the south-west coast of Fiordland at a depth of 25 km was felt from Stewart Island to Wellington. No damage was reported.
Two earthquakes centred in the Kermadec Islands region were felt widely in New Zealand. On 12 August, a magnitude 6.1 quake at a depth of 340 km was felt from Auckland to Invercargill, most strongly in Wellington, and on 12 October, a magnitude 6.2 quake at a depth of 150 km was felt from Auckland to Marlborough, mainly along the east coast of the North Island.
Christchurch activity continued at a low level in 2013, with three aftershocks of magnitude 4.6 being the largest recorded. The latest on 18 November caused a few instances of minor damage.
No significant swarm activity occurred during 2013.
The GeoNet Duty Team responded to a tsunami alert following the Santa Cruz Island (Solomon Islands) earthquake on 6 February. A marine threat was in place for some coastal area of New Zealand for around eight hours and maximum wave amplitudes of around 20 mm were measure at a number of tsunami gauge sites.
In 2013, there were four earthquakes of magnitude 6.0 or greater, and 66 of magnitude 5.0 to 5.9. About 2,500 earthquakes were reported felt.
The magnitude 6.2 quake struck at 1:07am and was located 125 km west of Tuatapere at a depth of 25 km. As of 9:30am we have received 234 felt reports from the public, most reporting little damage. This lack of damage is due to the earthquake occurring far enough away from the main populated areas of the region. There have already been several aftershocks from this quake, though most have been too small for people to feel.
Earthquakes in the Fiordland region are due to the collision of the Australian and Pacific plates. The mainshock early this morning was a large reverse faulting or thrust mechanism resulting from the Australian plate pushing (subducting) beneath the Pacific plate. Fiordland sits upon on the Pacific plate.
It is unlikely that this earthquake will have an effect on the Alpine Fault hundreds of kilometres further to the north.
Fiordland is one of New Zealand's more seismically active regions - due to the plate subduction - but it is also our most sparsely populated region, there have been a dozen large earthquakes (M6+) in the region in the previous twenty years, many that were much larger than the Canterbury earthquakes, but none have caused nearly as much damage due to the sparse population and people's distance from the quakes epicentres. These Fiordland quakes often also release their energy slightly slower, so that the shaking is less violent than an earthquake of equivalent magnitude elsewhere in the country. The motion is more "rolling" rather than "snapping" which will contribute to the lack of damage.
The largest quake Fiordland has experienced in recent times was the M7.8 Dusky Sound earthquake on 15th July 2009. That quake was New Zealand's largest since the 1931 Hawke's Bay earthquake, it occurred close to the coast and produced a small local tsunami. The quake early this morning was 70 km southwest of the Dusky Sound earthquake.
We've covered this ground before and made some small improvements along the way, but we are pretty pleased with this latest revision. You'll be used to seeing the location and magnitude jump around in the Quake History. We are now giving the magnitude a longer time to settle down before sending the alert to you. When I say "longer", we're only talking about 20 to 30 seconds. We think it's worth awaiting a wee while longer for something that's better!
This morning, at 9:44am, there was a magnitude 4.6 quake in Christchurch, 10 km to the east of the city. Our automatic system missed this quake, and we’re still investigating why this was. The earthquake was subsequently manually located by our duty seismologist.
The strong quake should have been easily recognised by the automatic system, as it registered on 77 of our seismographs around the country, as far north as Hamilton. In an hour since the quake GeoNet has received over 500 felt reports for this quake. Canterbury last experienced a quake of this size 8 months ago, on 19th January 2013.
Apart from missing this earthquake, the earthquake location system is functioning as normal, it has located over 70 quakes in the last 24 hours, including a quake 23 minutes before this missed Christchurch quake, and one 26 minutes later.
Earthquake activity in the Rotorua basin was typically quite high (1-200 events per year) but over the last decade or so this has declined to much lower levels. Since 2011 there has only been 26 local shallow earthquakes. However over night there was a bit of catching up. GeoNet has been able to locate 8 earthquakes, the largest was M 2.7 at 1.52 am and 257 people have reported feeling it. Many local residents have reported been woken and some have felt the smaller events.
To keep you updated with the ongoing research and details regarding Friday's Lake Grassmere earthquake, we've started a webpage in the Cook Strait Earthquakes sub-section of 'Our News'.
We'll be adding to this regularly over the coming weeks as we get new information, and as the sequence progresses.
You can find out more about the quake in our historic quakes section here.
The Cook Strait earthquakes page has further information including ongoing response and research on the Cook Strait and Lake Grassmere earthquakes and their aftershocks.
There will be an aftershock sequence to follow this earthquake, and we already have had one of magnitude 6 and several over magnitude 5. The latest revised forecast for a further magnitude 6 or greater, for the 24 hours from 12:00 pm 21 August, is 2.2%. See the full statistics table.
A deep magnitude 6.0 earthquake 800 km north-east of Whangarei at 4:16 pm (NZST) was earlier notified as two earthquakes of M4.8 and M6.1 in the Gisborne region. This is a similar situation to the occasions explained in "Tonga quake reported as two shakes off Tolaga Bay" and "Dilemmas of deep, distant earthquakes".
The New Zealand network is unable to locate earthquakes at this distance from the country quickly and accurately, but they do show up on our instruments and systems. Our duty officer is also alerted to them, and she checked them before removing them from the website as being incorrect and outside the New Zealand region we report on.