By studying the extremely dense human concentrations present at the opening of the San Fermín festival in Pamplona, researchers have uncovered a novel dynamic regime linked to the self-organization of compact crowds—an observation that could help prevent catastrophic crowd movements.
a) Photo of the crowd during the opening ceremony of the San Fermín festival (2019).
b) Aerial view of Plaza Consistorial with the analyzed area (dotted polygon) and observation points (orange dots). Scale: 33 feet (10 m).
c) Close-up view of the crowd 57 minutes and 15 seconds before the festival's start, showing the positions of participants' heads (green dots).
d) Close-up view 30 seconds before the festival's start, showing a significantly increased density.
e) Average crowd density over time before the start. The diamond indicates t = −57:15 (see c) and the star t = −00:30 (see d). The density increases slowly and steadily. The inset shows the radial distribution function g(r).
f) Local density maps at t = −57:15 (left) and t = −00:30 (right). White areas indicate obstructed views. Dense crowds are undoubtedly one of the most dangerous environments in our modern societies. The collective movements that emerge within them can indeed lead to uncontrolled displacements of groups of individuals totaling masses of several tons and, in the most tragic situations, result in significant human casualties due to trampling and suffocation.
However, modeling crowd dynamics primarily relies on interaction-based individual models, which, though effective for describing small groups, struggle to explain the complex behaviors of very dense crowds composed of thousands of people.
In a recent study, researchers from ENS Lyon, Université Lyon 1, and the University of Navarre (Spain) analyzed the movements of thousands of people gathered in Plaza Consistorial in Pamplona (Spain) during the opening ceremony of the San Fermín festival. They discovered that in confined spaces, dense crowds can self-organize into giant oscillators.
Without any external prompting, thousands of individuals spontaneously coordinate their movements to follow circular trajectories with a certain periodicity. Based on these observations and fundamental physical principles, the authors developed a mechanical model that accounts for the emergence of these collective movements.
This model reveals that extremely specific friction forces (called odd) induce, at high density, a collective phase transition that gives rise to chiral oscillations: very large fractions of the crowd coherently begin rotating in a random direction with a characteristic period of 20 seconds.
These results not only explain the experimental observations but also the fact that similar dynamics are observed during catastrophic events, such as the Love Parade in Duisburg (Germany) in 2010, where 21 people lost their lives.
Image of Plaza Consistorial in Pamplona (Spain). A crowd of over five thousand people gathers every year on July 6 at 12:00 PM to celebrate the opening of the San Fermín festival.
© Bartolo Lab, ENS Lyon.
The robustness of the experimental observations reported in this work and the minimal nature of their mechanical explanation open the door to a monitoring protocol to better anticipate catastrophic emergent behaviors in massive crowds. These results are published in the journal
Nature.
Reference:
Emergence of collective oscillations in massive human crowds,
François Gu, Benjamin Guiselin, Nicolas Bain, Iker Zuriguel & Denis Bartolo, Nature, published February 5, 2025.
Doi:
10.1038/s41586-024-08514-6
Open archive:
HAL