Traditionally, computational climate and weather system science has worked by splitting up a computer model of the Earth’s surface into a grid and calculating how components, such as bodies of land, air and water, and the Sun’s incoming heat, all interact. Central to our picture of what’s going on are the laws of thermodynamics – which govern how energy is moved around between bodies (such as the Sun, the Earth’s atmosphere and everything else!) and detailed maps of the Earth, but as with most science, one of the greatest leaps has been how scientists make careful simplifications to the big picture to make things easier to work out. This is especially important for complex systems, where small changes or what scientists call perturbations can dramatically affect the model’s prediction in a way that simply doesn’t reflect reality. Supercomputing has enhanced weather and climate research in number of ways: by including new interactions and data, modeling at a finer scale and considering many scenarios simultaneously.