method of characteristics

A first-order PDE looks intimidating because the unknown depends on several variables at once, but there is a beautiful trick that tames it: find the special curves through the domain along which the PDE collapses into an ordinary differential equation. Travel along one of these curves and the equation tells you simply how the solution changes as you go. These privileged paths are the characteristic curves, and following them is the method of characteristics.

Take a first-order equation like a(x, y) u_x + b(x, y) u_y = c(x, y, u). The left side is the directional derivative of u along the vector (a, b), so along any curve whose tangent is (a, b) — that is, a curve satisfying dx/dt = a, dy/dt = b — the change in u is governed by du/dt = c. The single PDE has become a system of ODEs: solve for the characteristic curves in the (x, y) plane, and along each one integrate du/dt = c to carry the solution forward from the data. Geometrically, the solution surface is woven out of these characteristic curves, each one launched from a point of the initial curve and carrying its value. For the simplest case, the transport equation u_t + c u_x = 0, the characteristics are the straight lines x - c t = const and u is simply constant along each — a profile that translates rigidly to the right at speed c.

The method of characteristics is the natural tool for first-order PDEs and for hyperbolic problems generally: traffic flow, gas dynamics, geometrical optics, and any conservation law. Its great honest lesson is what happens for nonlinear equations: when the characteristic speed depends on u itself, characteristics can cross, the solution surface folds over, and a smooth solution ceases to exist — a shock forms. The method does not merely solve the easy cases; it predicts exactly where and when classical solutions must break down, which is information no other elementary method gives you.