A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
A bifurcation occurs when the qualitative behavior of solutions to dy/dx = f(y, μ) changes as a parameter μ varies—typically when equilibria are created, destroyed, or collide. Bifurcation analysis reveals how system dynamics depend sensitively on parameters.
From autonomous equations and phase-line analysis, you know how to find equilibria of dy/dt = f(y) and classify them as stable or unstable by checking the sign of f'(y) at the equilibrium. You also know how to draw the phase line: a picture showing which intervals have solutions moving upward (f(y) > 0) or downward (f(y) < 0). Bifurcation theory asks what happens when the equation contains a parameter μ, so you have a family of equations dy/dt = f(y, μ), and you watch how the phase line changes as μ varies.
The simplest and most important example is the saddle-node bifurcation. Consider dy/dt = μ − y². When μ < 0, the equation f(y) = μ − y² = 0 has no real solutions — no equilibria, and all solutions move in one direction forever. At μ = 0, there is exactly one equilibrium at y = 0, but it is neither stable nor unstable in the usual sense (it is called a half-stable equilibrium). When μ > 0, two equilibria appear: y = +√μ (stable) and y = −√μ (unstable). At μ = 0, a stable and unstable equilibrium collide and annihilate each other as μ decreases — or, reading the other way, a stable-unstable pair is *born* as μ increases past 0. The value μ = 0 is the bifurcation point.
Other common bifurcation types include the transcritical bifurcation, where two equilibria exist for all μ but exchange stability as they pass through each other, and the pitchfork bifurcation, where one equilibrium splits into three at the bifurcation point (one losing stability while two stable ones are born). The pitchfork is common in symmetric systems — the classic example is a ball balanced on top of a curved surface, which is unstable but can tip stably to either side. The bifurcation diagram visualizes these changes: it plots equilibrium values y* against the parameter μ, using solid curves for stable equilibria and dashed curves for unstable ones. At a bifurcation point, the curves meet or branch.
Bifurcation analysis matters because real systems always have parameters — population models have birth and death rates, physical systems have temperature or pressure. Small changes in a parameter can cause sudden dramatic changes in long-term behavior — a population that was growing suddenly faces extinction, or a structure that was stable suddenly buckles. The bifurcation diagram tells you precisely where those critical thresholds lie and what happens at them, turning qualitative phase-line analysis from a snapshot at one parameter value into a complete map of how behavior depends on the parameter.