Warning, indulgent work content

As well as coming over for the food, culture and hockey, Telle and I also have to occasionally breeze past the university to do some work. Generally, we'll avoid talking about work, but some of you might be wondering what I actually do for a living, having spent almost a decade at university becoming "the" Dr., and the rest of you will just have to suck it up and find out anyway.

I now work in the field of complex systems, more commonly associated in the public psyche with chaos (for those keeping track my PhD was in non-equilibrium thermodynamics, but to be fair these aren't hard and fast divisions). There are a number of different ways of describing complex systems, but the way I like to think of a complex system is one that has a set of simple rules that tell you how it works, but when you actually put these rules into practice it exhibits unpredictable/unexpected behaviour.

For my opening work, I'm playing around with the complex Ginzburg-Landau equation. To describe the CGLE imagine a grid of paper, and at every intersection of the grid imagine there is an arrow. The value of the CGLE is given by the size of the arrow (the magnitude) and the direction the arrow points (the phase). As time goes by the value of the CGLE at each point changes based on its current value, and the difference between it and its neighbours. We can tweak the behaviour of the system based on two numbers, one that controls the contribution from its current value and one the controls the contribution from the difference in its neighbours. Just by playing around with these two values, we can radically change the behaviour of the system with time, but exactly what will happen isn't immediately clear before you simulate it (though that's not true now, because the system has been pretty heavily studied and you can actually get diagrams describing what it will do at various values).

Anyway, to test my program, I took two values described in the literature as generating a vortex glass, which is a state where you get whirlpools form in the system that repel each other and lock up the system into a series of whirlpool domains. I started with a random field (all the magnitudes and phases chosen randomly), let it evolve, and then after a while you get this (the color of each point corresponds to its phase).