What is a phantom traffic jam?

Above a critical density, a single driver's brake-tap propagates backward as a shockwave. Stop-and-go jams form and dissolve with no obstruction, accident, or bottleneck causing them.

Why does EDSA jam every rush hour?

EDSA's design capacity is about 2,400 vehicles per hour per lane, but peak demand reaches roughly 3,500. Demand exceeds capacity, pushing density past critical and collapsing throughput into stop-and-go.

EDSA · traffic flow

Jams from nothing. Rush hour on a highway that has too many cars.

Vehicles spawn on the left and drive right. Push the spawn rate past EDSA's critical density and watch stop-and-go waves form spontaneously — then ripple backward against the flow.

free flow

Density

—veh/km/ln

Avg speed

—km/h

Throughput

—veh/h/ln

Rush-hour controls

Spawn rate90 veh/min

lightEDSA peak →gridlock

Lanes5

25 (EDSA)8

Demand vs capacity—

below critical│ = capacity

How to use

Set the spawn rate and number of lanes, or hit a preset. Cars are colored by speed — green free-flowing down to red stopped. Below the critical density everything streams. Above it, jams nucleate on their own and crawl upstream.

The fundamental diagram

Flow is Q = density × speed. At low density speed is free, so flow rises with more cars. But speed falls roughly linearly once density climbs past ~25 veh/km/lane, and collapses near ~80. Past that point adding cars lowers total throughput — the road carries fewer vehicles per hour even though it's more crowded.

EDSA's design capacity is about 2,400 veh/h/lane (HCM range 1,900–2,400). Peak demand runs near 3,500. Demand exceeds capacity, so density is forced past critical and EDSA spends roughly four hours each rush in stop-and-go.

Phantom traffic jams

Above the critical density, one driver's brake-tap forces the car behind to brake harder, and the one behind that harder still. The disturbance propagates backward as a shockwave: a jam that forms and dissolves with no accident, bottleneck, or obstruction.

Sugiyama's 2008 experiment proved it — drivers told to hold a steady speed on a circular track spontaneously produced a stop-and-go wave that traveled around the loop. You'll see the same clusters appear here: dense knots of red that weren't there five seconds ago, drifting upstream.

Citations

Nagel & Schreckenberg (1992) — cellular automaton traffic model, J. Phys. I France 2, 2221.
Sugiyama et al. (2008) — empirical evidence for phantom jams, New J. Phys. 10, 033001.
HCM 2016 — Highway Capacity Manual, TRB. Lane capacity ~1900–2400 veh/h/ln.
JICA-MMDA Mega Manila Public Transport Study (2014) — EDSA demand/capacity gap.
Lighthill & Whitham (1955) — kinematic-wave model, Proc. R. Soc. A 229, 317.

Built by Joshua Chua · Fable rebuild · vanilla JS + Canvas