the Lorenz system
/ LOR-ents /
In 1963 the meteorologist Edward Lorenz, trying to capture the essentials of atmospheric convection, stripped a weather model down to three simple equations in three variables — and stumbled onto chaos. Re-running a forecast from a value he had rounded from six digits to three, he watched the new run diverge wildly from the old. From that accident grew the modern theory of chaos, and the Lorenz system became its founding example, the equation everyone meets first.
The Lorenz system is three coupled first-order ODEs: x' = sigma (y - x), y' = x (rho - z) - y, z' = x y - beta z, where x, y, z loosely track the intensity and temperature pattern of a convecting fluid, and sigma, rho, beta are parameters. With the classic values sigma = 10, beta = 8/3, rho = 28, the system has no stable fixed point and no stable cycle to fall into; instead every trajectory is drawn onto the Lorenz strange attractor, the two-winged butterfly. The motion loops some unpredictable number of times around one wing, then jumps to the other, never settling and never repeating, while remaining forever bounded in a finite region. It exhibits all the hallmarks: sensitive dependence, a positive Lyapunov exponent, a fractal attractor.
The Lorenz system matters far beyond meteorology because it was the first clean demonstration that a simple, fully deterministic, low-dimensional system of differential equations can be genuinely chaotic — overturning the comfortable assumption that simple equations have simple long-term behaviour. It is the standard testbed for every chaos concept (the butterfly effect is named after its attractor's shape) and a permanent caution: Lorenz's own lesson was that perfect equations plus imperfect initial data make long-range weather prediction fundamentally impossible, not merely hard.
Integrate the Lorenz equations at sigma = 10, rho = 28, beta = 8/3 from x = y = z = 1. The point spirals out on one butterfly wing, circles a few times, flips to the other wing, circles again, and keeps switching wings in a sequence no formula can shortcut — bounded forever, yet never tracing the same path twice.
The Lorenz trajectory loops one wing, then unpredictably switches to the other, forever — chaos in three simple equations.
The Lorenz equations are a drastic three-mode truncation of fluid convection, not a literal weather model — they capture the qualitative onset of chaos, not quantitative forecasts. And the system is chaotic only for suitable parameters (around rho = 28); for small rho it has a perfectly stable fixed point and no chaos at all.