Visible Earthquakes

At a conference in Knoxville I got to chatting with Gareth Funning about using some of my software to visualize earthquakes using InSAR data. Here is what wikipedia has to say about InSAR:

Interferometric synthetic aperture radar, abbreviated InSAR or IfSAR, is a radar technique used in geodesy and remote sensing. This geodetic method uses two or more synthetic aperture radar (SAR) images to generate maps of surface deformation or digital elevation, using differences in the phase of the waves returning to the satellite[1][2][3] or aircraft.

Did you read the part about surface deformation?! If a satellite flies over a piece of land before and after an earthquake occurs you can compare how much the ground has moved down to the millimeter. The data you get out of the satellite is called an interferogram and looks something like this:

Interferogram of the Dinar earthquake.

You can count up the rings or fringes to determine the amount of deformation in wavelengths; here each wavelength is 2.83cm.

What can you do with this data? Well, being scientists and all, we like to classify things like earthquakes so that we can learn about their mechanisms, structure, and life-cycle. Was it a normal fault, a thrust fault, strike-slip ...? If we know more about an earthquake that just happened we can compare to different areas and make predictions and inferences about if an earthquake is likely to happen in a different area we care about!

So, what we need is a game plan, how to take this (pretty) picture and make an educated and informed guess about the world. Here is the roadmap:

  1. use a simple model of the earth,
  2. simulate a model earthquake with known geometries, size etc.,
  3. estimate the interferogram that the known earthquake creates,
  4. compare the simulated data to the true data,
  5. contemplate deeper meanings, make a better guess, and repeat until happy.

So after talking with Gareth and exchanging some code for the simple model (step 1.) I mocked up a simple InSAR modelling package that allows you to iteratively make an informed guess about the earthquake that produced the interferogram. Here is a screenshot of the program:

On the left is the true interferogram, and on the right is the model. You can change the length, strike, dip, rake, depth, and amount of slip of the fault. Most of these you can do by graphically pulling around a few handles. You can see the model update in (near) real time and see the 3D fault block representation.

Why is this cool?

Try it Now!

The application right now is in alpha release, we record the submitted results at the moment, but don't show them to you. There are also some problems in Chrome, so I suggest trying it in Firefox if you have problems. Please leave comments!

https://earthquakes.aranzgeo.com/

If you need to brush up on your fault terminology:

Fault Lab

Visible Earthquakes in use

A screenshot (no audio):