Researchers test smartphones for crowdsourced earthquake warning

Seismogram being recorded by a seismograph at the Weston Observatory in Massachusetts, USA. Credit: Wikipedia

Smartphones and other personal electronic devices could, in regions where they are in widespread use, function as early warning systems for large earthquakes according to newly reported research. This technology could serve regions of the world that cannot afford higher quality, but more expensive, conventional earthquake early warning systems, or could contribute to those systems.

The study, led by scientists at the U.S. Geological Survey and published April 10 in the inaugural volume of the new AAAS journal Science Advances, found that the sensors in smartphones and similar devices could be used to build earthquake warning systems. Despite being less accurate than scientific-grade equipment, the GPS (Global Positioning System) receivers in a smartphone can detect the permanent ground movement (displacement) caused by fault motion in a large earthquake.

Using crowd-sourced observations from participating users' smartphones, earthquakes could be detected and analyzed, and customized earthquake warnings could be transmitted back to users. "Crowd-sourced alerting means that the community will benefit by data generated from the community," said Sarah Minson, USGS geophysicist and lead author of the study. Minson was a post-doctoral researcher at Caltech while working on this study.

Earthquake early warning systems detect the start of an earthquake and rapidly transmit warnings to people and automated systems before they experience shaking at their location. While much of the world's population is susceptible to damaging earthquakes, EEW systems are currently operating in only a few regions around the globe, including Japan and Mexico. "Most of the world does not receive earthquake warnings mainly due to the cost of building the necessary scientific monitoring networks," said USGS geophysicist and project lead Benjamin Brooks.

Researchers tested the feasibility of crowd-sourced EEW with a simulation of a hypothetical magnitude 7 earthquake, and with real data from the 2011 magnitude 9 Tohoku-oki, Japan earthquake. The results show that crowd-sourced EEW could be achieved with only a tiny percentage of people in a given area contributing information from their smartphones. For example, if phones from fewer than 5000 people in a large metropolitan area responded, the earthquake could be detected and analyzed fast enough to issue a warning to areas farther away before the onset of strong shaking. "The speed of an electronic warning travels faster than the earthquake shaking does," explained Craig Glennie, a report author and professor at the University of Houston.

The authors found that the sensors in smartphones and similar devices could be used to issue earthquake warnings for earthquakes of approximately magnitude 7 or larger, but not for smaller, yet potentially damaging earthquakes. Comprehensive EEW requires a dense network of scientific instruments. Scientific-grade EEW, such as the U.S. Geological Survey's ShakeAlert system that is currently being implemented on the west coast of the United States, will be able to help minimize the impact of earthquakes over a wide range of magnitudes. However, in many parts of the world where there are insufficient resources to build and maintain scientific networks, but consumer electronics are increasingly common, crowd-sourced EEW has significant potential.

"The U.S. earthquake early is being built on our high-quality scientific earthquake networks, but crowd-sourced approaches can augment our system and have real potential to make warnings possible in places that don't have high-quality networks," said Douglas Given, USGS coordinator of the ShakeAlert Earthquake Early Warning System. The U.S. Agency for International Development has already agreed to fund a pilot project, in collaboration with the Chilean Centro Sismologico Nacional, to test a pilot hybrid earthquake warning system comprising stand-alone smartphone sensors and scientific-grade sensors along the Chilean coast.

"The use of mobile phone fleets as a distributed sensor network—and the statistical insight that many imprecise instruments can contribute to the creation of more precise measurements—has broad applicability including great potential to benefit communities where there isn't an existing network of scientific instruments," said Bob Iannucci of Carnegie Mellon University, Silicon Valley.

"Thirty years ago it took months to assemble a crude picture of the deformations from an earthquake. This new technology promises to provide a near-instantaneous picture with much greater resolution," said Thomas Heaton, a coauthor of the study and professor of Engineering Seismology at Caltech.

"Crowd-sourced data are less precise, but for larger earthquakes that cause large shifts in the ground surface, they contain enough information to detect that an has occurred, information necessary for ," said study co-author Susan Owen of NASA's Jet Propulsion Laboratory, Pasadena, California.

Explore further

3D satellite, GPS earthquake maps isolate impacts in real time

More information: Crowdsourced earthquake early warning, Science Advances,
Journal information: Science Advances

Citation: Researchers test smartphones for crowdsourced earthquake warning (2015, April 10) retrieved 20 September 2019 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments

Apr 10, 2015
I don't know why it wouldn't work for smaller ones... I have a "seismograph" app on my phone (for entertainment purposes only). It's just a graphical display of the X/Y/Z accelerometer values, X-axis on the plot is time (I think 100hz sample rate) and Y-axis is G's. I have used it to display EQs in the 2's and 3's... Then again, I'm also just a few miles from them, so you would be less likely to have a large sample size of phones that would have picked it up.

It's been easy for me to play with apps like this because in 2014 there were 863 EQs <10mi from my house, 96 of those were 3.0+ (172 and 22 respectively for 2015 YTD). When they're really close (<2 miles), you can feel them even in the low 2's, and a mid-3 six miles away can wake you up... a low 4 that close can crack sheetrock, and with so many of them, slabs and exterior masonry too. We had a 4.3 two days ago... Welcome to Oklahoma; where the oil industry comes first. 2014 total for the state was 5,417 with 586 3.0+.

Apr 11, 2015
Hmmm..... It sounds like making simple earthquake samplers that would be sensitive enough to pick up small quakes would be a good idea. If they could make them use wifi or even a cable hookup, the scientist could probably talk many business's into simply keeping one on the premises for nothing other that the small publicity gained by being part of a free sensor network. If the boxes do nothing but send data when a quake occurs, I don't believe most people would have a problem with one being hooked up to the network. I would hook one up to my home system if it would help, no problem.

Apr 11, 2015
How do you separate normal jiggles from walking or anything else from earthquakes? Seems to be to many disconnects in this. Perhaps a seismograph sending signals would be better?

Apr 14, 2015
@gkam: I suspect if you're in the middle of walking or in a car, it would be difficult. However, many phones are sitting still on a desk, in a purse on a table, in a dock, etc. or nearly still while sitting in your pocket while you're sitting down. There are characteristics that would indicate an EQ vs. other movement. It could also only transmit the data if it was still, then suddenly had movement that was characteristic of an EQ.

As a follow-up to what I said about detecting small ones; I guess it isn't as important to detect small ones, since a) they don't travel as far, b) the damage they do is not USUALLY catastrophic or life endangering (although, our 16' tall and 6' wide "stacked-look" stone chimney/fireplace in my living room scares me a little whenever we have an EQ). Regarding how far they can travel. evidently, in the US, EQs east of rocky mountains propagate further than EQs in the west. I know there have been some low 4's in Oklahoma that were felt in Nebraska.

Apr 14, 2015
MP3, that seems to be because our ground is already fractured here, and the disturbance travels poorly in fractured strata.

Here, we have learned to tell the P from the S waves. The P waves are compression waves, and will make the house creak, before the S waves get here to do the real damage. Early warning systems use P waves as a signal.

Apr 14, 2015
the good thing is, most the ones we get here in OK are pretty short. Usually there is a boom/rumbling sound just before the shaking. The boom is followed by a brief (maybe 2-15 seconds) high frequency shuddering, so you get .25-.5 seconds of "warning. But the initial boom/jolt is almost as if a big dog plopped down up against your bed or the couch... then the shaking is more like a dog on your bed scratching itself. The longest shaking lasted for around 20-30 seconds. My belief is that the quantity of them is playing more of a role in damage than the strength. An occasional 3-4 wouldn't be much to worry about I'm sure. But with as many as we have, I think it can weaken wood framing. Not the wood itself of course, but the joints. For example, take cheap furniture using cam locks and move it from from one place in your house to another. Then move it again, and again, and again. Eventually, it will be "wobbly" for lack of a better term...

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more