The U.S. Geological Survey prides itself on being first with earthquake alerts thanks to its nearly 2,000 real time sensors.
But since most of those are in the United States, a temblor can go unnoticed especially in remote parts of the world. So to close that information gap, the USGS has increasingly turned to Twitter and its hundreds of millions of users for help – effectively creating an ad hoc detection system.
Together, USGS seismologist Paul Earle and software developer Michelle Guy were tasked with looking at how Twitter data could be used for earthquake detection and verification. By using Twitter’s Public API, they used the same time series event detection method they use when detecting earthquakes, according to a blog post this week from Twitter Communication Manager Eliane Ellis.
In a 2011 paper on using Twitter, Earle, Guy and another author found that Twitter messages came through within tens of seconds after the initial shaking and that a time series of Tweets with the word earthquake could identify when the worst shaking occurred.
“The detections are generally caused by widely felt events that are of more immediate interest than those with no human impact. The detections are also fast; about 75% occur within two minutes of the origin time,” the authors wrote in the study that appeared in the Annals of Geophysics. “This is considerably faster than seismographic detections in poorly instrumented regions of the world. The tweets triggering the detections also provided very short first-impression narratives from people who experienced the shaking.”
Digging further, they found that many earthquake-related Tweets were short as in “earthquake?” so they started filtering out tweets with more than seven words. They also found that few people shared links or the exact size of a quake, so they filtered out any Tweets sharing a link or a number. Together, this filtered stream proved to be very significant at determining when earthquakes occurred globally.
During a visit to the USGS offices in Colorado to meet up with Earle and Guy, Ellis said three earthquakes happened. Using Twitter data, Ellis said the USGS system was able to pick up on an aftershock in Chile within one minute and 20 seconds – and it only took 14 Tweets from the filtered stream to trigger an email alert.
The other two earthquakes, off Easter Island and Indonesia, weren’t picked up because they were not widely felt, she noted. That is typical of the 70 or so quakes that happen daily, since many occur deep in the ocean or in uninhabited areas.
By using Twitter data, earthquakes can be detected faster - helping trigger an alert often in less than two minutes. Beyond the speed, the USGS can mine Twitter for words in other languages - terremoto, for example, was used in Chile to indicate a bigger quake - to spot an earthquake and even use the silence on Twitter to conclude that initial reports of a quake in a populated area were unfounded.
Next, the USGS team says that they want to determine if they can drop Twitter data based detections into seismic algorithms, and if that can speed up alerts even more.