North Canterbury, New Zealand Earthquake
On 14 November 2016 at just after midnight local time a major earthquake struck in the North Canterbury region of New Zealand. This isn’t surprising as the country sits on the boundary of the Australian and Pacific tectonic plates, part of the Earth’s crust, and it has a long history of earthquake activity.
Some of the impacts of this 7.8 magnitude quake make amazing reading. Scientists have estimated that the two biggest land masses of New Zealand, the North and South Islands moved closer together by 2 metres! Around the other side of the world the Large Hadron Collider (LHC) picked up the earthquake waves as they distorted the 27km in circumference particle accelerator ring by 0.1mm:
The LHC is undoubtedly the world’s most expensive (and unintended) earthquake detector. However they need to adjust the alignment of the sub-atomic particle beams even for movements like this.
So why am I talking about earthquakes? Well, inspired in part by @DarthPathos I wanted to experiment with using SAS to analyse earthquake data. Fortunately there is a great website run by Geonet, a public organisation in New Zealand that freely provides this data: http://www.geonet.org.nz/
I extracted earthquake data into a CSV file starting from 5 days prior to the big earthquake, to 5 days after. I then tried the SAS SGPLOT procedure to get this:
North Canterbury New Zealand Earthquake: Magnitude 7.8 Depth 15km
As you can see the magnitude 7.8 earthquake stands out like a sore thumb. Leading up to the big event there are very few quakes happening and those that do happen are quite small. You wouldn’t normally feel these unless they were very close by.
However contrast that with the quake storm that follows. There are literally hundreds of them, with dozens exceeding magnitude 4 or 5. These you would definitely feel even if you were some distance away.
I then decided to enhance the SGPLOT to tell more of a story.
North Canterbury New Zealand Earthquake: Magnitude 7.8 Depth 15km
I used the DROPLINE statement to identify the big shake and I’ve overlaid the depths of the quakes using the COLORRESPONSE option. It is interesting to note that most of the aftershocks are quite shallow – less than 30km. Generally speaking the shallower the quake, the more violent the shaking so the quake depth is an important measure. What is also interesting about the aftershocks is the trend in the magnitudes. Initially the magnitudes are high but they decay over the days and weeks following in an exponential trend.
Here is the complete program for this graph:
proc sgplot data= WORK.EARTHQUAKES ;
title "North Canterbury New Zealand Earthquake: Magnitude 7.8 Depth 15km";
where depth <=50;
scatter x = origin_dt y = magnitude / group = Quakegroup colorresponse = depth colormodel = (red green blue black);
dropline x = "13NOV2016:11:02:56"dt y = 7.82 / dropto = both lineattrs=(color=red pattern=dot) ;
run;
Fortunately the 7.8 earthquake happened in sparsely-populated area and at a time when most people were at home so there were only 2 deaths and a small number of injuries. Here is a map from SAS VA that shows the area. This was very quick and easy to do. I loaded the earthquake dataset into the VA LASR library. I then used the earthquake ID to define a custom geographic location using latitude and longitude. I then created a Geo Bubble Map report and defined the size of the dot to show the earthquake magnitude and the colour to show the depth. This screenshot really doesn't do justice to the amazing level of detail. I've also only shown the magnitude 5 or higher quakes otherwise the map would just be obliterated with dots.
On a personal note, I live in the city of Wellington which is about 200km from the epicentre of the 7.8 earthquake. There was significant damage in the city even at that distance including a lot of buildings that are now having to be demolished or to have major repairs. The building I work in needs some minor repair work so I could be working offsite for a month or two! Meanwhile the aftershocks are continuing and we are still feeling them on most days.
So there you have it. A real life experience that motivated me to use SAS to explain what living through earthquakes is like.
