Earthquakes are avalanches. So are crackling fires, Rice Krispies in a bowl of milk and crumpling pieces of paper. The sounds we hear as crackling are a result of avalanche behavior. Avalanches, as physicists understand them, aren’t made of snow: they are responses to stress.
In the Laboratory of Atomic and Solid State Physics, Prof. James Sethna, physics, and postdoctoral research associate Stefanos Papanikolaou study theories that predict the random behavior of avalanches.
“Avalanches occur in physical systems that are slowly put under stress, like when you crumple a piece of paper,” explained Sethna. “As your fingers slowly push the paper, the system [paper] responds to the stress with abrupt events with a variety of sizes. Those events are called avalanches.”
Piles of sand and snow, Sethna noted, were once thought to exhibit avalanche behavior, and hence, the name we now associate with snow tumbles, but technically, in the physics realm, they do not qualify. Even popcorn kernels popping are not considered avalanche behavior because each sound’s intensity is about equal.
This is where statistics are used in physics: the key to a true avalanche is the distribution of responses. Some sounds are very small, some are larger and some are really big. This mix of intensities defines the distinctive noise we hear as crackling, which results from avalanche behavior that makes noise. Avalanches can also be seen in seismograph data resulting from tectonic plates shifting.
Papanikolaou and Sethna published field-advancing findings on avalanche behavior in the journal Nature Physics last month.
Their work presented two major breakthroughs. First, Papanikolaou used high quality data free from usual distortions seen in avalanches generated in magnets. He interpreted data from collaborators in Brazil who created a super thin magnet and generated crackling data. Tiny disordered crystals inside magnets shift to align their poles when placed into a coil, an effect known as Barkhausen noise. The thinness of the study’s magnet minimized distortions usually seen when using thicker magnets. But because the magnet was so thin (1 μm), the size of the avalanche data was small. Papanikolaou had to apply special data analysis methods to filter out background noise from the weak signals.
The second major breakthrough of their work occurred when the theoretical physicists were able to predict random behavior. “We finally achieved the level where the theory of avalanches quantitatively agrees with experiments,” Papanikolaou said.
Avalanches are scale invariant, which means the general shape of the noise is roughly the same regardless of the scale used to look at them. Whether zooming in or out, the spread of avalanche peak heights looks about the same. Much like fractals, avalanches are self-similar; the bigger picture is composed of smaller representations of the same shape.
This concept of symmetrical, predictable behavior — known as universality — is at the core of physics, but providing observed data to support these theories is challenging. Sethna and Papanikolaou have broken ground in theoretical physics by predicting more complicated behaviors than just so-called power laws that link intensity size with frequency. Their theoretical work builds on a complex mathematical approach called the renormalization group, advanced by famed former Cornell physicists Kenneth G. Wilson and Michael E. Fisher, and which won Wilson a Nobel Prize in 1982.
Sethna and Papanikolaou’s work on predicting behavior from avalanche data has real-life applications. Earthquakes are one example of how looking at a smaller scaled data —seismic events lower on the Richter scale — might allow scientists to predict larger avalanche events, like earthquakes.
If you are interested in learning more about this research, you can hear the difference between crackling and imposter noises recorded by Sethna’s group at http://simscience.org/crackling.