Samuel Williams Saturday, 05 March 2011

There are many theories regarding the Christchurch earthquakes — but few look at the local environment for possible explanations. I am not a geologist; however after discussing the various possibilities I thought we came up with an interesting possible explanation.

The recent earthquakes we have been experiencing in Christchurch are potentially due to a reduction of water in underground aquifers. Because there is less water, this causes a reduction of pressure in the aquifers which then causes them to collapse or shift.

To explain this idea, I've made some diagrams. Here is an example of the water table under the Canterbury planes (obviously depth not a linear scale). At some point the pressure of the water water and sediments must have an equilibrium:

Water and Earth in balance.

However, if water input is reduced (and thus the water table drops), then it may be possible for these structures to collapse into the void created by an absence of water. As the major structures collapse, the minor sediments are pushed up through the gaps due to the large pressure exerted on the remaining water supply.

Earth collapses and pushes up sediments.

Essentially, the structure of the underground aquifers is supported by the pressure exerted by large amounts of water — and that the use of water for farming and irrigation has caused these underground structures to become empty. This in turn causes massive internal collapse. The remaining water and sand is then pushed up through the ground, as the weight of earth is now pressing down on top of it with no resistance.

The evidence for this is the vast amounts of water and sand that has been pushed up from underground during several of the last major earthquakes. Even if this phenomenon is not responsible for the main earthquake, I wonder if it contributed significantly to the destruction of the roads and underground structures around Christchurch.

Are we destroying the natural Christchurch aquifers by drawing too much water? It would be great to have some geologists comment on this theory.


You are not the only one thinking the “unthinkable”

I have spoken to several geologists here in Canterbury, all of whom agree this theory has merit, and deserves investigation, however to seek funding for such research would be impossible due to the political and economic interests involved in irrigation here in Canterbury.

I got an interesting response via email:

Last Saturday (2/7/11) there was a fascinating article in The Press on the first Canterbury farmer (near Pendarves 1968) to use bore irrigation to convert from sheep to cropping… he was initially considered to be wasting his money by neighbours and government farm advisers! How quickly they changed their tune.

I have considered this a possibility all along and spoken to geologists, some who agree and some who disagree with the hypothesis that water extraction/aquifer collapse may have caused/triggered the quakes.

The difficulty would be in proving the hypothesis, as it may or may not be correct.

Significant political and economic forces would seek to silence any such research, and scientists who still have a job, mostly want to keep it!

The same week that GNS convinced the Canterbury University Geology Department not to release the findings from their collaboration with Canadian experts (who brought vital equipment to measure faults), the government announced its $420 million funding for a huge new irrigation scheme between the Rakaia and the Waimakariri! … coincidence?

Its important to remember that irrigation triggered quakes is only a theory, a possibility, but I suspect it is one the government and business do NOT want explored for the very obvious reason it would cast a shadow over their desire to further develop irrigation in Canterbury … Big money is involved.

If your theory was true then you would have regular small-scale, local earthquakes around aquifier extraction sites, and not significant region-wide tremors. Unless Christchurch is above a cavern, the water table isn’t that fluid.

Interesting article. However I think a more likely explanation would be the fact that we are on a contental plate boundary. Which is converging causing the formation of the southern alps.4mm growth per year for those mountains is a huge ammount of strain. You have to remember that he quakes are at bedrock depth whcih is not supported by pore pressue, also, the gravels are not supported by pore pressure, they are of a high sand and silt matix and stacked on top of each other. Just go down to the rakaia gorge and take a look. Those gravel cliffs are what the aquifers look like.

I am a postgraduate Geologist at the University of Canterbury, specifically I am hydrologist/geochemist and I have had some involvement with the seismic survey, which was a joint project between the University of Canterbury and the University of Calgary.

First, in response to a comment which does concern me, I have not heard of any pressure from GNS on the University of Canterbury and the University of Calgary not to publish data from the seismic survey. I would very much like to see the source for that statement as it is disconcerting. As far as I am aware the data from these surveys are still being processed, seismic surveys produce a huge volume of information that requires a fair amount of processing to remove background noise and interference which takes a considerable amount of time and computing power. I think the preliminary results should be available by the end of July or start of August and will link any publications as they come out of it if you’re interested.

While the sediment does compress when you extract groundwater you do not see an upward movement of fine grained materials and water from this action. The amount of sand and water ejected during the earthquakes is the result of a process known as liquefaction, which has done considerable damage to the roads, homes and infrastructures around the city. Liquefaction occurs when the shaking caused by the earthquake essentially removes the stress holding the saturated fine layers down allowing them to compress and flow like a liquid which then flows to the surface. Wikipedia: Soil Liquefaction has a more detailed explanation.

The areas of liquefaction have more to do with the subsurface geology more than any lowering of the groundwater tables. Christchurch is built over floodplains where the Waimakariri, Avon and Heathcote rivers have meandered and moved over time depositing sands, silts and gavels over much of the city. The old river channels contain the majority of the silt and sand, the fine particles, and it’s in these old river channels, that the majority of liquefaction occurs. If you look at the government deemed red zone, it roughly parallels the current river channels, as the areas around the current rivers have the most recently buried channels.

Christchurch overlies and relies upon the Christchurch Artesian Aquifer System, which is comprised of a number of distinct aquifers under separated by confining layers (impermeable layers of mud and clay in this case). The greater Canterbury plains, comprises the Canterbury Plains Aquifer, (we are really creative in naming), which is a large unconfined aquifer which contributes a large part of the water to the Canterbury plains. This water in turn is the primary water source for the Heathcote and Avon rivers, with groundwater derived from a recharge zone in the Waimakariri feeding the Avon, and possibly ‘deeper’ groundwater feeding the Heathcote.

To answer your question at the end of your blog, currently there has not been what would be classified as a dramatic drop in the water table due to irrigation, this is monitored very closely by ECAN and NIWA, however the conversion to dairy is something that needs to be monitored as this is a much more intensive style of farming and will result in more nitrates and phosphates entering the groundwater which is harmful for the environment, and is a primary cause of eutrophication, and in large enough doses human health. The impact of the conversion to dairy on the Canterbury Plains and the Christchurch Artesian Aquifer is the center of a lot of debate in both the scientific and local community and is a subject I feel I do not yet know enough about to explain well.

An interesting paper on the subject of the effects of groundwater removal is a 2007 paper which examined the long term changes of a aquifer as it was over pumped then allowed to recharge over the course of 30 years.

Another good blog would be Dr Quigs; Mark Quigley is a geomorphologist specializing in active tectonics, a professor and researcher at the University of Canterbury.

  • Sorry about the wall of text just realize how much I’ve written but feel it was necessary. I cut out a significant chunk of text on the effects of lowering of groundwater tables and its impact on the surface but will be more than happy to talk about them with you. Sorry about any grammatical mistakes I may have made as well, spent a lot of time writing this but still ran out of time before I could proofread.

Mike Hutton asks for the source of where GNS asked University of Canterbury not to release data from their seismic survey with Calgary (Canada) scientists): Data on city faults out next month (originally published in The Christchurch Press).

Interesting that the data still hasn’t been released nearly TWO months later…I understand The Press have requested it under the Official Information Act, and are still waiting, but maybe Mr Hutton can investigate why the information was suppressed by the GNS?

Next, the gist of the irrigation causing quakes argument has nothing to do with Liquefaction…that is a red herring, the issue here is that the rather unique geological structure that is the Canterbury plains, is essentially a huge lens shaped formation of gravel deposits, within which vast underground ‘rivers’ or aquifers exist. Underneath the gravel deposits, are the solid perhaps fractured rock layers that bear against the main alpine fault. The theory is that the massive extraction of water from the aquifers has sufficiently changed the pressure balance within the underlying structures to trigger the series fo movements we have all experienced. The volume of water extracted in central Canturbury annually exceeds five Benmore dams…A litre of Milk takes 10000 litres of water to produce…these are serious volumes and weights and depleted aquifers.

The Greendale fault, where it all started, is in the Red zone of over allocation of irrigation water, long ago recognized by Ecan as endangering the serious degradation of the underlying aquifers.

Surely the scientific response could be more thorough and convincing than that so far given?

I just got a reply that there was a press release based off the preliminary data on June 3rd, It had completely slipped under my radar, and I apologize for not double checking the GNS press releases.

First off thank you for your source; I don’t think there was any pressure from GNS not to present the preliminary results but probably a decision to wait until all the data had been collected or that processing the data was taking longer than expected, it does happen. However as I was not involved with these decisions I can only speculate. I do know that at time of writing of that article, May 12th, the seismic crew had only run lines along New Brighton beach, Barbados Street, and Highfield road over the Greendale fault. Over the course of May, basically from May 12th-30th, three more seismic lines were run over the Canterbury plains including a 16 Km line on Robinson’s road. During this time I know a number of other geophysics projects were being conducted, I know an electromagnetic survey was run with a helicopter while we were on Robinson’s road in mid May.

The post about liquefaction was not intended to be a red herring in any way but rather to explain why large volumes of water and sand were pushed up from the ground in the Christchurch, and why it is not the result of structural collapsed due to a lowered water table.
I had left this out of my initial post due to length but will include it now. There are surface changes that occur when groundwater table is lowered dramatically; however an increase in seismic activity has never been documented as result of this. There have been a large number of studies on aquifer deformations due to the effects of massive groundwater removal. Generally the most common issue is subsidence, or the lowering of the land surface relative to sea level, which occurs when the groundwater table is lowered and the sediments start to compress. This poses problems to infrastructure, increased risk of aquifer exhaustion, higher risk of flooding and seawater incursion. Again, there is no evidence that a lowered water table or extraction could trigger earthquakes.

Upon reading your description and looking at the diagrams in this post, there appears to be a fundamental misunderstanding of what an aquifer is and how water flows through aquifers. While there is a lot of talk about flow and how water moves through the ground, this transport is very different than how rivers that flow at the surface. A good analogy would be to fill a glass with marbles then fill the glass with water. Water occupies what is called the pore space between sediments, or the gaps in the marbles in this analogy. The ability for water to flow in the ground is a measurement of hydraulic conductivity, which is related to how connected the pore space is to one another. The larger your sediment size, and therefore the larger the pore space, the easier it is for water to move through. The aquifer doesn’t exist as rivers between rocks and sediments; the aquifer is the water flowing through the pore space of the rocks and sediments.

The ground beneath the central Canterbury plains is about 600 m thick of gravel alluvial fans with various lenses of sand silts and clays, again old river channels usually, laid down by braided rivers over time. A good modern analog near Christchurch for the environment which laid down these materials would be the Waimakariri. Underlying these gravels is the greywacke, a hard sandstone that makes up the Torellese group which forms the basement rock for New Zealand. Christchurch and the greater Canterbury plains do have large and obviously active faults as a result of the tectonic plate boundary through the country. Not all the stress of the plate motion is taken up by the alpine fault, and it is taken up by the various faults in the region, including the Greendale fault.

While there is a significant amount of water being extracted from the Canterbury Plains Aquifer, there is also considerable recharge to the aquifer. Groundwater allocation limits are usually a maximum of 50% of the annual groundwater recharge. This is based off of annual assessments of the groundwater resources and land-surface recharge, the percent of water used in irrigation, and rainfall that soaks back into the soils. The amount allocated is not necessarily the amount being used, for instance in Waipara a significant amount of water allocated is not being used.

Without a doubt the Canterbury Plains Aquifer is under stress and being degraded, this is term which indicates both amount of water and water quality. In terms of water quantity the aquifer is not being depleted. Well levels have dropped but there still is a considerable amount of water still in the aquifer, however this does have impacts huge impacts on the environment especially the springs fed rivers in the region. Even though I am highly skeptical of the figure 10,000 liters of water per 1 liter of milk, cows are extremely land intensive and require far more water and fertilizer than sheep. This fertilizer enters the groundwater and ultimately winds up in our streams and lakes, Lakes Ellismere is extremely polluted in terms of nutrient pollution and further land use intensification will only exacerbate this. But I digress in terms of numbers well levels have dropped from 2002-2008, but increased in 2009, wells do fluctuated with changes in climate, land use and both short term and long term weather patterns. These are issues of water quality and detracting from discussion of a hydrological trigger for the September 4th earthquake.

Some links /sources of further information:
ECAN’s state of the Canterbury water resource reports 2009-2001
ECAN’s Groundwater Allocation Limits: land-based recharge estimates September 2004
Ecan report: Age and source of Canterbury plains groundwater Report No. U02/30 September 2002

**again Massive wall of text, sorry. If anyone interested in any books on hydrology or environmental science I can give a few suggestions. Also if I have made grammatical errors it’s because I just realized I spend a large amount of time writing this and its now very late.

The volume and flow of the Canterbury aquifer has been dramatically reduced in recent years due to massive extraction for irrigation. The ground temperature below the plains increases by 30-35 degrees C every 1000 metres,(Russ Van Dissen GNS science the Press 3/8/11) ie boiling point at just 3 km. Vast pressures are involved in this giant seam of steam and rock and referred seismic stress. The cooling effect of the aquifers on the underlying bedrock has been greatly reduced, perhaps triggering the September quake, which in turn triggered the later quakes(Jarg Pettinga GNS science the Press 3/8/11). Has the unique location of the Canterbury aquifer over a seismic and volcanic ‘hotspot’ combined to make irrigation a very risky business in this region?

Over 8,000 earthquakes all in the underground catchment area of the Waimakariri River from Darfield to Ellesmire to Kaiapoi. We also know of aquifers deep off the coast of Kaiapoi. How deep do the aquifers under the Canterbury Plains and Christchurch go…15ks or more? …the depth of some of the earthquakes under Christchurch? Is irrigation/water extraction depleting these aquifers creating subsidance, earthquake faultline cracking and a giant sinkhole?…. Destruction…Heartache….Responsibility?…Negligence?….Insurance?…..Culpability?

Check out the theory of ‘Hydroseismicity’ in ‘New Scientist’( 14 January, 2012) “Rapid changes in the water table….can suddenly change the fluid pressure in these faults – and that might trigger earthquakes”. (p.35)

How deep do the aquifers under the Canterbury Plains and Christchurch go? They go no deeper than circa 500m below surface. They can’t possibly go to 15km depth as one poster states…. the maximum thickness of the Canterbury Basin is around 6000m and this occurs in the offshore portion of the basin beyond the eastern limits of the aquifers.

All of this information is readily accessible via any decent search engine.

“Essentially, the structure of the underground aquifers is supported by the pressure exerted by large amounts of water — and that the use of water for farming and irrigation has caused these underground structures to become empty. This in turn causes massive internal collapse. The remaining water and sand is then pushed up through the ground, as the weight of earth is now pressing down on top of it with no resistance.”

Your diagram shows a “soup” of water with sedimentary clasts that are not in contact with each other. This is not the case. Unless a reservoir is grossly overpressured (which the Canterbury Aquifers aren’t) the clasts within a reservoir are in contact with, and supporting each other and the water lies in the pore spaces between the clasts. The structure of the reservoir rock/sediment is supported by the clasts making up the sediment/sedimentary rock not the fluid between the clasts. Consequently if you reduce the pressure of the fluid in the pore spaces the structure of the reservoir is not going to collapse. All over the world there are oil and gas reservoirs that are severely depleted due to years of production….there are no cases of such depleted reservoirs collapsing as they become depleted and reservoir pressure decreases.

Check out James Syvitski and Stephanie Higgins, University of Colorado, ‘Swamped -River deltas all over the world are sinking’( New Scientist, 1 December , 2012. pp40-43) “If fluids like oil and water are removed at high rates, the overlying land can sink. This is exactly what happened in Thailand. From the 1950s to the 1980s, water was sucked up at such a rate that buildings in Bangkok began to sag and crack.”… “ Fluid injection has also been shown to produce hundreds of microearthquakes, and over long periods of time, it can cause earthquakes of magnitude 4 or larger. Groundwater will slowly seep back into collapsed aquifers even without injection, but it does not “reinflate” all the tiny spaces it used to occupy”.

I found your article whilst looking for information on this very subject–namely, could reduced aquifer levels under Christchurch be a causative or contributing factor in the earthquake situation there. I wondered about this, after reading an article suggesting that California may experience increased earthquake activity as a result of very low aquifer levels (CA aquifers are massively overdrawn, as opposed to what I understand is the situation in Chch where we are told the aquifer levels are not actually dropped significantly). Here’s an article on the situation in California.

That article also mentions seismic activity in Spain (2011) some scientists hypothesised may be linked to lowering aquifer levels.

Leave a comment

Please note, comments must be formatted using Markdown. Links can be enclosed in angle brackets, e.g. <>.