Since it opened in 1995, the National Drought Mitigation Center at the University of Nebraska-Lincoln has collected data on drought and drought impacts across the U.S. In 2005, that data collection was expanded to include news reports of drought conditions. The data is used to compile the U.S. Drought Monitor every Thursday, a tool used by climate scientists, meteorologists, farmers and countless others who rely on the bounty of the skies.
Kelly Helm Smith, the Assistant Director of the Drought Mitigation Center wondered if social media could add to the story and the accuracy of the drought monitor, so she studied two years’ worth of drought tweets and developed a method to measure the rate of drought tweets over time in a given geography.
“Sometimes people can detect what’s going on a little bit quicker than the instrumental record, especially with something like drought where it’s not just about how little precipitation you got, but it’s also about all sorts of conditions are that are contributing to that,” Smith told the AgriTalk Radio Show. “People can certainly talk about that pinch that they might be feeling before it might be showing up in other ways.”
Smith collected two years worth of tweets on #drought, #drought17, #drought18 and so on to see if they could serve as an early warning system for oncoming drought.
“What interests me the most is when you get farmers and ranchers tweeting about their actual field conditions, or taking a little video clip off the back of their tractor showing how dusty it is,” she said. “Sometimes that really does provide new information.”
The key, she says, is for those who post about drought and soil moisture conditions to include location data with the tweet.
The Twitter data will be added in with the other data points for the drought monitor in order to provide another level of resolution on the developing drought picture. Smith also distributes a map of drought tweet information.
Smith is now working to dig deeper into tweets that don’t necessarily mention drought but could still provide clues to researchers.
“I would really like to be able to do some deeper searches, not using the drought hashtags, but looking at some of the terms that people use in their tweets, like, you know, crunchy corn or cracks,” Smith explained. “I actually have a list going and have different words that people tend to use to describe dry conditions. And so what if we did a search for all of those terms and then look at what are we learning from that? So getting away from #drought to just conversations about ‘boy sure was dry this week.’ How do we find those tweets?”
The pilot study was published in the Bulletin of the American Meteorological Society.
