What the Fitbit Saw During That 4.4 Quake

Map of MyShake phone locations

Map shows locations of Android phones that alerted during the Berkeley quake on January 4. (Berkeley Seismological Lab)

Some of the most useful new research on earthquakes might be coming from — well, you, via your smartphone and wearable activity tracker.

As a magnitude 4.4 quake jolted people out of their beds around the Bay Area last week, some of those folks had smartphones that were already at work sending data to scientists via the MyShake app developed at the UC Berkeley Seismological Lab.

According to Richard Allen, who heads the lab, instantaneous data came in from 59 phones armed with the app, which uses built-in motion sensors to detect the strength of the shaking and transmit the data to scientists. They can use that data to produce useful maps of the temblor’s reach and impact. The app responds so quickly to a quake’s first vibrations, that at some point in the next few years, plans are to make it part of a phone-based warning system that can alert users to quakes seconds before the shaking starts where they are.

Allen was impressed that readings came in from phones farther than 60 miles from the epicenter.

“The data is really clean for this event,” wrote Allen in an email, “which encourages us that we can use this for research purposes, and eventually earthquake early warning.”

The MyShake app is currently available only for Android phones. But potentially useful data is beginning to come in from other sources that weren’t developed with earthquakes in mind.

Of Fitbits and Fitfulness

Developers of the Fitbit series of wearable activity trackers say that when the quake struck, they also got a burst of data. Since the devices also track sleep patterns, they inferred from that data that about 4-in-10 of their Bay Area users were awakened by the quake, which struck at 2:39 a.m. Thursday near the historic Claremont Hotel in Berkeley.

The company declined to say how many Bay Area users it has, but a Fitbit spokeswoman told KQED that the numbers are based on “a representative sample of tens of thousands of aggregated and anonymous Fitbit user data.”

Fitbit quake graph
Spike in graph shows people waking up as the Jan. 4 M4.4 quake shook the Bay Area, as registered by Fitbit devices. Click to enlarge. (Fitbit)

Analysts at the San Francisco-based company could virtually map the epicenter of the quake by looking at the proportion of Fitbit users awakened in various locations. As one would expect, Berkeley users had the biggest wake-up call, with the percentage of wakeful customers leaping from 8 percent to 52 percent as soon as the Hayward Fault started slipping. (The fact that even 8 percent were already awake might imply that Fitbit has a loyal following among insomniacs.) The farther from the epicenter, the fewer Fitbit users were jarred awake — just 18 percent in San Jose, for example.

It’s still unclear what value this kind of data might have for seismic research.

“Smartphone data has been quite useful for us at the [lab] so far, and other techniques such as using fiber optic cables are showing promise as well for seismology,” notes Jennifer Strauss, who is in charge of outreach at the Berkeley lab. “So, it would be very interesting to explore how other crowdsourced data can be used for earthquake science.”

Fitbit analysts also noted that even relatively benign earthquakes can be night-wreckers, as it took an hour for most awakened users to return to “normal sleep,” and 90 minutes before those rocked out of bed achieved the REM stage of sleep.

What the Fitbit Saw During That 4.4 Quake 16 January,2018Craig Miller

  • Laura Morland

    This is fantastic news, as my Fitbit monitors my sleep every night. On the other hand, I didn’t realize that “Big Fitbit Brother” was watching my sleep — I thought my stats were just between my tracker and my devices. It’s a brave new world we’re in.

  • BlueGreen

    Second the ‘big brother’ comment, and how quick they are to show that they are gathering this formation. I’m sure it’s in those ridiculous terms of service agreements that basically render the device inoperable if you don’t’ agree. As with all this information, it’s cute to gather it and come up with neato uses like plotting the epicenter (already done by USGS) but how well are they going to secure my information from those who would use it for their own selfish/ possibly dangerous reasons. Knowing someone is asleep is a powerful piece of data!

Sponsored by

Become a KQED sponsor