Previous: A World Within Our World: Hang Sơn Đoòng | Weird Places
Next: The Link Between Zebra Stripes and Sand Dunes | Natural Patterns



View count:198,330
Last sync:2023-11-15 03:15
Devastating earthquakes happen every year, and it's difficult to predict when they will happen. But they do follow one mathematical pattern known as the Devil's staircase.

Hosted by: Hank Green

SciShow has a spinoff podcast! It's called SciShow Tangents. Check it out at
Support SciShow by becoming a patron on Patreon:
Huge thanks go to the following Patreon supporters for helping us keep SciShow free for everyone forever:

Kevin Bealer, Jacob, Katie Marie Magnone, D.A. Noe, Charles Southerland, Eric Jensen, Christopher R Boucher, Alex Hackman, Matt Curls, Adam Brainard, Jeffrey McKishen, Scott Satovsky Jr, James Knight, Sam Buck, Chris Peters, Kevin Carpentier, Patrick D. Ashmore, Piya Shedden, Sam Lutfi, Charles George, Christoph Schwanke, Greg
Looking for SciShow elsewhere on the internet?


 (Fast Fact)  Science isn't just about explaining the natural world, it's also about taking those explanations and using them to make predictions. Predictions can save lives. But there are also natural events that so far, we can not just forecast.

 New Section (0:17)

For example, devastating earthquakes happen every year, and even though scientists broadly understand why. They've never successfully predicted one. But as erratic as earthquakes seem, they do appear to follow one odd mathematical pattern known as "The Devil Staircase".  Which is a thing that is about as bizarre as it sounds. And yet the combination of geology and abstract math could point the way to better earthquakes predictions. 

 New Section (0:44)

Powerful earthquakes are relatively rare, on average, there are about 16 each year. And traditionally we've understood that they happen as tension builds along a fault in the  Earth's crust and then gives way. However we never know when that will happen. Today scientist are able to predict the average number of events that will happen during a given period, but not when or where they will happen.

Which might help government budget for disaster relief but does not help government's issue like evacuation warnings. 

 New Section (1:15)

Earthquakes might not be as random as they seem though. In a 2020 study a team of researches noticed these events loosely follow a well-known pattern called "The devil staircase". It's more properly called the cantor function after the German mathematician George Cantor, who discovered it in 1883.  Almost any point you pick, the Cantor function is flat, but there are also spots where it rises and a bunch of steps that are all really close together. And it works sort of like like a fractal.

If you zoom in, on any given part. The Cantor function looks basically the  same, with long periods of inactivity interrupted by short burst of activity, Like a really terrible staircase. 

 New Section (1:55)

 But actually turns out to be pretty useful for representing some things in nature. Especially the behavior of certain highly connected systems, like brain cells.  For instance scientists have used The Cantor Function to explore why certain brain activity might trigger a seizure. It's useful because small changes to these systems happen all the time and mostly have very little effect, tha's the flat part of the function. But when something big does happen, it can often trigger more big events that happened one after the other.

Like a series of quick steps.

 New Section (2:25)

The researchers observed patterns of earthquakes in several regions around the world and found that a number of them seemed to follow the Devil's Staircase. Including in the western's US, Australia, and off the cost of Algeria. Basically within certain regions tectonic activity in one part of the crust can trigger activity nearby. The analysis suggest that in these areas, short burst of strong quake activity happen more often that you'd expect from traditional modeling.

That's likely because traditional models focus on the tension at a single fault but in these areas, the time between burst seems related to the amount of strain on broader regions of the Earth's crust.

 New Section (3:07)

So a model that treats each fault individually seems to be missing the bigger picture. Unfortunately, all that interconnectedness leads to cycles that are less predictable than we'd hoped. So it's not a huge step forward in predicting future events, but recognizing that powerful earthquakes happen in burst could helps us understand what actually happens at a fault. And that could pay off big in the future, after all it doesn't seem likely that earthquakes are truly random events.

They're just processes that are so complicated, that we haven't figure it out what factors really matter.

 New Section (3:41)

In the case of earthquakes, the devil really is in the details. Thanks for watching this episode of SciShow, and if you liked it you might enjoy our episode about whether or not we are overdue for a mega quake. You can watch that, right after this.