Method of Characteristics for First-Order PDEs

How do one predict where a shock wave will form during supersonic flight? How do one model traffic jams on a highway? How does a tsunami change shape as it approaches shore? All of these questions have their answer hidden in characteristic curves, special lines along which a complex PDE collapses into a simple ODE.

• **Aerodynamics**: supersonic aircraft generate shock waves (sonic booms); the method of characteristics predicts their shape and location
• **Traffic theory**: the Lighthill-Whitham model treats vehicle density as u_t + f(u)_x = 0; traffic jams are shock waves in the flow
• **Oceanography**: shallow-water equations are quasilinear PDEs; characteristics give the propagation speed of a tsunami

Предварительные знания

• Nonlinear ODEs and Qualitative Analysis
• ODEs and PDEs in Machine Learning

Characteristics: curves along which a PDE becomes an ODE

Unreal Engine 5 and Blender Cycles solve the radiative transfer PDE via the method of characteristics: each ray is one characteristic curve. The PDE over 3D space collapses to an ODE along the ray, making real-time path tracing at 60+ fps possible on modern GPUs.

A characteristic is a curve (x(s), t(s)) in the (x, t)-plane, parametrized by s, such that the change in the solution u along it is governed by an ordinary differential equation. Rather than solving the equation everywhere at once, we trace the evolution along these curves.

Think of the surface u(x, t) over the (x, t)-plane. The PDE constrains the slope of that surface. A characteristic is a curve in the (x, t)-plane along which we 'slide' over the surface u without violating the equation. Key property: along a characteristic, du/ds is given by a simple formula that depends only on values along the same curve.

The algorithm: prescribe initial data on a curve (e.g., u(x, 0) = u0(x)), launch characteristics from each point of the initial curve, track how u is transported along each characteristic, then reassemble the solution in (x, t)-coordinates.

Linear transport equation: a traveling wave

The simplest case is the **linear transport equation**: u_t + c*u_x = 0, where c is a constant. Physically, u(x, t) is the concentration of a substance carried by a flow at speed c.

u(x, t) = u0(x - ct) means: to find the value at position x at time t, look at the initial profile at x - ct. The wave 'runs' to the right at speed c. Examples: a sound wave at small amplitude, spreading of a dye in a river, a light pulse in a homogeneous medium.

In the linear case the characteristics are parallel straight lines. They never intersect, so the solution exists and is unique for all t > 0. This is the ideal situation. In the nonlinear case things get more complicated.

Quasilinear case: the inviscid Burgers equation

In a quasilinear PDE the coefficients depend on the solution itself. The canonical example is the **inviscid Burgers equation**: u_t + u*u_x = 0. The transport speed is c = u, the solution itself! This makes the characteristics nonlinear.

Higher values of u move faster. If the initial profile is decreasing somewhere (u0'(x) < 0), the 'crest' catches up with the 'trough' ahead of it. Characteristics from different points converge, and the wave steepens. This models: gas dynamics (shock waves), shallow-water waves, traffic flow (vehicle density), nonlinear acoustics.

Shock formation: when characteristics cross

When characteristics intersect, the classical solution ceases to exist: u would be multi-valued at the crossing point. Physically this corresponds to a **shock wave**: a sharp discontinuity in the profile of u. Describing discontinuous solutions requires generalized theory.

After t_break a smooth classical solution no longer exists. One introduces a **weak solution**, where the PDE is satisfied in an integral sense rather than pointwise. Weak solutions may not be unique; an entropy condition is imposed (physical: a shock compresses characteristics, not expands them). Applications: aerodynamics (supersonic flight), blast waves, shallow-water tsunamis, traffic jams.

The method of characteristics is more than a mathematical trick. It reveals the physics: information in hyperbolic PDEs propagates along characteristics at finite speed. This is the fundamental distinction from parabolic equations (diffusion), where perturbations are felt everywhere instantaneously.

Key ideas

• **Method of characteristics** reduces a first-order PDE to a system of ODEs: dx/ds = a, dt/ds = b, du/ds = c
• **Linear transport equation** u_t + c*u_x = 0 has the exact solution u(x, t) = u0(x - ct): a traveling wave that preserves its shape
• **Burgers equation** u_t + u*u_x = 0: each characteristic carries constant u, but speeds differ, so characteristics can intersect
• **Shock wave** forms at t_break = -1/min(du0/dx); after that, weak solutions and the Rankine-Hugoniot condition are required

Related topics

The method of characteristics is the foundation of hyperbolic PDE theory:

• Second-order PDEs: hyperbolic type — The wave equation is also solved via characteristics, which become cones in spacetime
• First-order systems of ODEs — The characteristic system is a system of ODEs; fluency with solution methods is essential
• Numerical methods for PDEs — Finite-volume methods and Godunov schemes use characteristics explicitly

Вопросы для размышления

• Why are characteristics always parallel in the linear transport equation but not in Burgers?
• Physical meaning of the entropy condition: why is an 'expanding discontinuity' (characteristics flowing out) unphysical?
• In traffic flow, u is vehicle density and f(u) = u*v(u) is flux. How does the method of characteristics explain the formation of jams?

Связанные уроки

• calc-17-multivariable