Where Wave Expansion Fails on Real Roads
Traffic waves are a useful approximation, but only on uniform stretches of road. Real networks break the assumption the moment intersections, signals, lane drops, or merges interrupt the free flow that the wave equation requires.
A complete interactive classroom, not just a preview.
Start when you are ready to enter this Stage's 8 scenes and explore, respond, and learn as you go.
Where does the kinematic wave-expansion idea break down on real road networks?
A clean grille model predicts traffic congestion spreading like a circular wave — but drivers have seen it spread very differently.
If a wave model were right, jams should ripple outward in clean arcs; on real streets you see jams appear, vanish, and reappear in surprising places.
Side-by-side animation of an idealized circular shockwave versus a city-grid simulation that shows jams spawned, dragged, and erased by intersections and signals.
The wave idea breaks down exactly where roads change — at intersections, signals, merges, and bottlenecks — because the kinematic-wave equation ignores these discrete events.
If traffic really behaves like a compressible fluid, a jam should spread outward in a smooth circle from where it started, slowing everyone behind it equally.
- Macroscopic fundamental diagram of cities
- Connected/automated vehicle control strategies
- Route choice and demand modeling
- Pedestrian or transit dynamics
- 01The Beautiful Wave PictureslideQuestion
Introduce the kinematic-wave idea: traffic density obeys a conservation law, so jams should propagate as shockwaves, just like sound or floods. Frame the puzzle: then why does real congestion look so jagged?
- Conservation of vehicles on a road gives a PDE
- LWR (Lighthill–Whitham–Richards) predicts smooth shockwaves
- If jams were true waves, they should fan out in clean arcs
- 02Commit to a HypothesisquizPrediction
Ask the learner to commit before seeing any road evidence.
- Pick the feature most likely to break the wave picture
- 03Idealized Wave vs. Grid SimulationinteractiveEvidence
Let the learner watch two traffic scenarios side by side: a single long highway link obeying LWR, and a real grid with signals and intersections. Spawn a small perturbation and observe how each disturbance spreads.
- Highway: smooth, roughly circular shockwave
- Grid: jam freezes at a signal, restarts downstream
- Compare front speed in both cases
- Toggle signal timing to see jams 'stick' differently
- 04Four Network DiscontinuitiesslideEvidence
Show, with diagrams, the four places where the wave picture visibly fails: signalized intersections, merging on-ramps, lane drops, and at-grade junctions where two streams cross.
- Signals turn a steady queue into a pulsed wave
- On-ramps inject a continuous source term the PDE ignores
- Lane drops force a capacity drop, not a smooth density change
- Cross-street traffic breaks one-directional flow
- 05Why the Equation BreaksslideExplanation
Explain that LWR assumes a single, continuous flow with a smooth fundamental diagram. Every network discontinuity introduces a boundary condition, a capacity change, or a discrete event that the PDE does not contain — so waves get reflected, absorbed, or regenerated instead of propagating cleanly.
- LWR assumes continuous x, t, density, and flow
- Signals = boundary conditions, not part of the PDE
- Merges = source terms the conservation law omits
- Lane drops = capacity discontinuities
- 06How Long Must a Road Be?interactiveBoundary
Let the learner shrink the link between two intersections and watch the wave regime collapse. Find the length below which no clean shockwave forms.
- Sweep link length from 500 m down to 50 m
- Watch shockwave identification (sharp rear-front) fail
- Identify the minimum length where LWR is useful
- 07Apply It to a New SituationinteractiveTransfer
Give the learner a ring road with five on-ramps. Ask whether the wave model can predict where jams will appear at 8 a.m. based on a single 6 a.m. downstream incident, and let them test by toggling ramp metering.
- Identify which on-ramp would break a clean wave
- Predict jam location with and without ramp metering
- Generalize: any recurring source term breaks the model
- 08Answer: Where the Wave Idea BreaksslideResolution
Resolve the driving question directly and place the wave model in its proper scope: useful as an approximation on uniform links, misleading on real networks.
- Wave model = correct on long, uniform links
- Breaks at every intersection, signal, ramp, lane drop, and merge
- Real networks need network extensions (e.g., link-transmission, cell-transmission, or junction models)
- Use waves for back-of-envelope forecasts, not for signalized or merged systems
Discussion threads for a Stage aren't available yet.
