Drop, cover and tweet. As that becomes common procedure after a tremor, the U.S. Geological Survey is using Twitter to detect earthquakes around the world.
In a blog post on Wednesday, Twitter communications manager Elaine Ellis detailed how the USGS has been using the social media network to keep tabs on earthquakes.
Seismologist Paul Earle and software developer Michelle Guy teamed up to create an algorithm that could filter through tweets and quickly determine if an earthquake had taken place.
The trick was eliminating tweets that were longer than seven words or included links or numbers. All of those were likely to be people tweeting facts or sharing news of an earthquake -- not firsthand reports from people who actually felt the ground shaking.
The ideal tweet looked something like this:
Using that method, the USGS is able to detect an earthquake in under two minutes, something that could be useful in the areas not covered by the agency's approximately 2,000 sensors.
"At first, the USGS staff was a bit skeptical that Twitter could be used as a detection system for earthquakes," Ellis wrote, "but when they looked into it, they were surprised at the effectiveness of Twitter data for detection."
