California is notorious for its earthquakes, but a stunning new discovery reveals for the first time just how much we've underestimated its omnipresent earth-shaking potential.
By the time you just about finish reading this story, in fact, Southern California will probably have experienced another quake – based off a new, unprecedented deep dive into 10 years' worth of seismic data, which isolated almost 2 million 'hidden' tremors in the region that scientists had never identified before.
For decades, scientists suspected these invisible, smaller quakes existed, but had no way of singling them out from other random vibrations created by things like vehicle traffic, construction projects, weather events, and more.
"It's not that we didn't know these small earthquakes were occurring," says geophysicist Zachary Ross from Caltech.
"The problem is that they can be very difficult to spot amid all of the noise."
Awareness of the prevalence of these quieter quakes in California dates back to the 1940s, when scientists discovered that smaller tremors outnumbered larger quakes by a factor of 10 for every drop of a single unit in magnitude.
But knowing something is there is not the same thing as measuring it.
To actually discern the smaller earthquakes amidst the constant low-level din of vibrations that seismometers pick up, Ross and his team used a technique called template matching, which looks for the virtually unique seismic signature each fault produces.
Basically, whether an earthquake at a given location is a major event or a small temblor almost nobody notices, the jagged vibrations detected by seismometers look the same.
"We look for something that has a near identical waveform," Ross told Science News.
In theory, then, if you have a recording of a big quake that happened at a certain place, you can pore through seismic data from the same location looking for the same crests and valleys at a smaller magnitude. If you find matching jags in the data, you're looking at another earthquake, just on a much smaller scale.
"Now we can automate it and search exhaustively through the full waveform archive to find signals of very small earthquakes previously hidden in the noise," says one of the team, seismologist Daniel Trugman from Los Alamos National Laboratory.
Template matching in seismology is not actually new, but nobody had ever done it on a large scale before, until now, mostly due to the computational requirements.
But with the help of an array of 200 computer graphics processing units, the researchers crunched 10 years' worth of seismic data (2008 to 2017) recorded by the 550 seismic monitoring stations that make up the Southern California Seismic Network.
After about three straight months of computer processing time – in which algorithms sifted through the vibrations to identify any potential quake matches – the epic search isolated over 1.8 million earthquakes in total: almost 10 times more quakes than we previously were aware of in Southern California.
The new data suggest earthquakes in the region are about 10 times more frequent than we knew, with around 495 taking place in the region every day, or about one every 3 minutes.
Any way you look at it, that's a massive discovery, but the biggest takeaway isn't just what the researchers say is the most comprehensive earthquake catalogue to date – but what the hundreds of thousands of previously hidden tremors can tell us about seismic activity in the region.
"In the laboratory, we see small events as precursors to big slip events, but we don't see this consistently in the real world," Trugman says.
"Now we've discovered quakes previously discounted as noise and learned more about their behaviour. If we can identify these sequences as foreshocks in real time, we can predict the big one."
That might sound like a huge claim, but the more we understand about how big quakes relate to little quakes – and how foreshocks relate to aftershocks – the closer scientists will be to understanding where we sit on earthquake timelines.
In the future, the team says these approaches could be used in other regions, and that advances in machine learning will make it possible to penetrate ever more inscrutable seismic squiggles.
"The extraordinary detail resolved in this type of catalogue will facilitate a new generation of analyses of earthquakes and faults," the researchers write in their paper.
"As seismologists detect increasingly smaller earthquakes, the average time between observed events will continue to decrease, revealing more complex dynamical behaviour."
Others say the most exciting applications might not just be predicting earthquakes.
The new quake catalogue – publicly available here – potentially stands to offer much more, some think, as it may provide evidence of other hidden forms of seismic phenomena that earthquake researchers couldn't previously make out.
"It's just like if a new telescope comes along and its magnification is 10 times greater," seismologist Greg Beroza from Stanford University, who wasn't involved in the study, told Scientific American.
"Suddenly we see all of these stars and planets that previously we didn't see."
The findings are reported in Science.