Share this
by Richard Leeke on 05 January 2011
I was at the Tableau data visualisation conference in Seattle when the Christchurch quake struck NZ in September, so I couldn't resist getting hold of some earthquake data from GNS and seeing what it looked like.
Ever since I got home I've been meaning to write this up for the blog, but I got busy as soon as I got back and never got around to it. But inspired by the boxing day aftershocks, plus some visual presentation ideas I got from this posting on Robert Mundigl's Clearly and Simply site, I finally got around to finishing it off.
The data visualisation below uses Tableau's free "Public" service, which allows interactive visualisations of publically available data to be hosted on Tableau's public servers. Anyone can create these visualisations, using the free version of the Tableau visualisation tool, available here.
The GNS site has details of around half a million New Zealand earthquakes, which is more than is allowed on Tableau Public, so I filtered down to the last 40 years and only included magnitude 3 and above, which just sneaks in under the 100,000 limit.
I have included several tabs in the visualisation. The main one allows you to explore the earthquakes in the vicinity of a given location, identified by a postcode. I've also included a couple of other tabs showing summary views of the maximum magnitude and maximum depth of quakes experienced over different parts of the country.
Enter a valid NZ postcode and a radius in km to display earthquakes near that location. You can also filter the display by date range, magnitude and depth.
As a starter, here are a few interesting postcodes:
Darfield: 7541
Edgecumbe: 3160
Wairoto (Fiordland): 9691
See the instructions tab for more details of how to navigate the page and also for suggestions of things to look at around these and a couple of other locations.
The "Magnitude" tab shows the maximum magnitude quakes experienced in different parts of the country, by dividing the country up into a grid. Adjust the resolution of the grid to see more or less detail.
This view also lets you explore different ways of visualising magnitude. The Richter scale (or more properly the Moment Magnitude scale, which is what is used these days) is a logarithmic scale, so comparing earthquakes visually with something proportional to the magnitude does not really convey relative size of quakes very effectively.
In this view, the magnitude is represented by both the colour and the size of the marks. But representing magnitude by size on a map like this can be misleading. Should the area of the mark be determined by the magnitude (which makes little distinction between a magnitude 3 which is not even felt and a magnitude 7 causing major destruction)? Or should it be a metric which conveys the power of the quake more realistically, such as amplitude or energy released?
Either way, the impression given is that the quake in some way "covered" the area of the mark, whereas in reality the area over which it was felt and the extent of the damage caused is quite different and depends on numerous other factors. Compare the impression of the relative sizes of quakes given by setting the mark size to magnitude, amplitude and energy. Which is the most appropriate?
The "Depth" tab shows the maximum depths of quakes experienced over different parts of the country, with the deepest one around Taranaki and up north.
Enjoy!
Acknowledgents
NZ earthquake data comes from GeoNet and is freely available thanks to the GeoNet project, sponsored by EQC, GNS Science and LINZ.
NZ postcode data comes from the GeoNames database, which has over 8 million international placenames available for free download.
The approach to finding earthquakes in the vicinity of a given location is based on Robert Mundigl's post refered to above.
The maths for dividing the earth's surface up into a uniform grid using latitude and longitude (which I used in the magnitude and depth tabs) comes from Robert Morton of Tableau.
See the "About this viz" tab for more details of each of these sources.
