Sometimes science is all about uncovering hidden patterns. When Daniel Zitterbart of the Alfred Wegener Institute in Germany spent some time in Antarctica researching seismology several years ago, he noticed something curious about the local Emperor penguins: the males tended to huddle together in very dense formations, often thousands at a time. It reminded him of cell dynamics --so much so, that he decided to videotape the huddled penguins and study the time-lapsed footage. His latest findings can be found in a paper published today in the New Journal of Physics.

There's a very good reason penguins huddle. It's frickin' cold in Antarctica -- temperatures can plummet to -50 degrees Celsius with high winds topping 200 km/hour -- and the creatures are just trying to stay warm, especially during breeding season. (Male penguins are in charge of incubating eggs during the winter months). The denser the huddle, the warmer they will be-- at least those lucky enough to position themselves deep inside the huddle, as opposed to shivering on the outer edges.

A couple of years ago Zitterbart published his initial findings: the penguins don't stand completely still, they actually move every minute or so, and when they do, all the surrounding penguins move with them -- again, because it's cold, y'all, and they're following the warmth. That movement basically constitutes a traveling wave, rippling through the huddle -- but Zitterbart wanted to learn more specifics about how such traveling waves propagate and what triggers them in the first place.

That's when the University of Erlangen-Nuremberg's Richard Gerum suggested they look at mathematical models designed for traffic flow. It's common for scientists to describe traffic flow in terms of phase transitions, such as how water can move from a freely flowing gas, to a slightly more viscous (but still flowing) liquid to freezing into an icy solid. The cars on a freeway are like the molecules in H20: the more cars there are, the more densely packed they are, the less freely they can flow, and the more likely it is traffic jams (the solid phase) will form.

But as I wrote in a 2011 blog post, there are critical nuances to that basic description. Specifically, in 1998 a physicist named Boris Kerner found a pattern of self-organization lurking in years of data collected from traffic moving along German highways. To wit:

[Kerner] developed a model that essentially broke traffic into three basic categories: freely flowing, jammed (solid state), and a bizarre intermediate state called synchronized flow, in which densely packed “car molecules” move in unison, like members of a marching band. When this happens — when all the cars are traveling at close to the same average speed because of the vehicle density on the roadway — they become highly dependent on one another. A physicist might compare the relationship to the correlated motion of electrons in metals, which gives rise to weird phenomena like superconductivity. Highly correlated traffic means that a tiny perturbation — a single driver braking unexpectedly — will send little ripples of corresponding slowdowns through the entire chain of cars behind him/her.

Those tiny ripple effects are very similar to the traveling waves Zitterbart observed in his huddled Emperor penguins; the penguins stop and start much like heavy traffic inches along the road -- except while it's irritating for drivers, for the penguins, this stop-and-go movement is a good thing, since that's how they maintain their huddles and keep warm. The waves also merge and form even larger huddles if they are triggered sufficiently close together in time. Check it out:

Zitterbart and Gerum's modified model recreates the positions, movements and interactions of each penguin in a huddle and tracks the huddle's movements over time. They found that any given penguin only needs to step roughly 2 centimeters in any direction, and its immediate neighbors will do the same to keep as close to that first penguin as possible.

And any penguin can trigger a traveling wave propagating in any direction, provided that initial 2 cm gap is there (a "threshold distance"). Why 2 centimeters? Well, according to Zitterbart, it's twice as thick as a penguin's insulating layer of feathers. There's basically a "sweet spot" where the huddle is as densely packed as possible without crushing their natural feathery insulation.

He admits he was surprised that any penguin can trigger a wave, since he'd initially expected the trigger would come from a penguin lurking on the edges of the huddle trying to push its way in. Still a mystery is why any individual penguin starts to move in the first place. "Right now we speculate that the penguins might use this technique to be able to rotate their eggs -- which they can't with their beak, [so] it would make sense for a penguin to start movement even if he is right inside the huddle," Zitterbart told me in an email. Or maybe they just need to shift their weight from time to time; it's can't be too comfy standing there, shivering on the ice, with thousands of your penguin brethren pressing in.

"Ultimately we want to create a formula to predict the state of a penguin huddle depending on the time of year, the number of penguins, and how cold it is," Zitterbart told New Scientist. That's because for penguins, a traffic jam dynamic is what enables them to survive in harsh conditions -- and with climate change models predicting various extinction scenarios, Zitterbart thinks now is the time to study the creatures. They're truly unique. "In contrast to most other collective dynamics systems in macrobiology (bird flocks, fish schools, sheep herds), penguins stay in a highly ordered phase when resting," he told me. "Most other systems rely on a moving system to maintain ordered phases."


Bando M. et al. (1995) "Dynamical model of traffic congestion and numerical simulation," Physical Review E 51(2): 1035-1042.

Gerum, R.C. et al. (2013) "Origin of traveling waves in an emperor penguin huddle," New Journal of Physics 15: 125022.

Gilbert, C. et al. (2006) "Huddling behavior in emperor penguins: dynamics of huddling," Physiology & Behavior 88: 479-488.

Helbing, D. (2001) "Traffic and related self-driven many-particle systems," Reviews of Modern Physics 73.

Kerner, Boris S. (1998) "Experimental Features of Self-Organization in Traffic Flow," Physical Review Letters 81: 3797-3400.

Waters, A., Blanchette F., and Kim, A.D. (2012) "Modeling huddling penguins," PLoS ONE 7: e50277.

Zitterbart, D.P. et al. (2011) "Coordinated movements prevent jamming in an Emperor penguin huddle," PLoS ONE 6: e20260.