pattern formation
Why does a uniform stew of chemicals organize itself into stripes? Why do sand dunes ripple, clouds form rows, and animal coats carry spots? In each case a system that starts out featureless and symmetric spontaneously develops structure on a definite length scale, with no template imposed from outside. Pattern formation is the study of how and when this happens — how order emerges from a uniform state purely through the internal dynamics of a (usually nonlinear) PDE.
The mechanism, in PDE terms, is a symmetry-breaking instability. You start from a spatially uniform steady state of an evolution equation. You then ask: if I add a tiny ripple of wavenumber k, does it grow or decay? Linearizing the equation around the uniform state gives a growth rate sigma(k) for each k — a dispersion relation. If sigma(k) < 0 for all k, every ripple decays and the uniform state is stable (boring). But if sigma(k) > 0 for some band of k around a preferred wavenumber k*, perturbations at that scale grow, and the most-amplified mode k* sets the spacing of the emerging pattern. The Turing instability in reaction-diffusion systems is the classic route, but the same story drives convection rolls (Rayleigh-Benard), the Swift-Hohenberg equation, and buckling. As the unstable modes grow, the nonlinearity eventually arrests the growth and selects a saturated pattern — stripes, hexagons, spots — whose details depend on the nonlinear terms.
Why does it matter? Pattern formation explains how complex spatial structure can arise without a blueprint, which is foundational for developmental biology, materials science, geomorphology and chemistry. The deep lesson for a PDE learner is that nonlinearity is essential: linear equations can only amplify or damp modes, but it takes nonlinear terms to STOP the growth and lock in a stable, finite-amplitude pattern. Honest caveat: predicting which pattern (stripes vs spots) is genuinely hard and depends delicately on the nonlinear coefficients, the geometry, and boundary conditions — linear theory gives you the length scale, not the final picture.
Heat a thin layer of fluid from below. Below a critical temperature gradient the fluid is still (conduction); cross it and the motionless state becomes unstable to a band of wavenumbers, and the fluid spontaneously organizes into regular convection rolls of a definite width (Rayleigh-Benard convection). The roll spacing is set by the wavenumber whose growth rate first turns positive.
A uniform state goes unstable at a preferred scale and structure appears.
Linear stability analysis tells you the length scale of the emerging pattern (the fastest-growing wavenumber), but it cannot tell you the final amplitude or whether you get stripes, spots or hexagons — that is decided by the nonlinear terms, which is exactly why these are hard, nonlinear problems.