Researchers at Penn State University have suggested drawing on social media resources to aid flood relief efforts, based on analysis of the Colorado floods which occurred in 2013.
“FEMA (the Federal Emergency Management Agency), the Red Cross and other response agencies use social media now to disseminate relevant information to the general public,” said Guido Cervone, associate director of the Penn State’s Institute for CyberScience. “We have seen here that there is potential to use social media data from community members to help identify hotspots in need of aid, especially when it is paired with remote sensing imagery of the area.”
Response teams often have to rely solely on data and imagery which is sent to them via satellite. However technological this sounds, there is a significant downside.
Elena Sava, a geography graduate student studying at Penn State, states that “publicly available satellite imagery for a location isn’t always available in a timely manner — sometimes it can take days before it becomes available. Our research focused on identifying data in non-traditional data streams that can prove mission critical for specific areas where there might be damage. We wanted to see if social media could help filling the gaps in the satellite data.”
And as it turns out, it can. Twitter and Flickr users uploaded real-time images of the floods and flood damage to the respective sites. Combining this method with remote sensing, the relief teams are more accurately able to pinpoint impacted areas and send relief.
The Colorado floods were unexpected, and dealt a tough blow to the surrounding areas. Forty-three centimetres of rain fell over the course of nine days – which is normally the amount that would fall in the area over the space of one year. Ten thousand people were evacuated from affected areas by officials, and rescue teams had to help a few thousand people out of the floods.
The upside to this devastating flood in such a metropolitan setting was the sheer amount of people who took to Twitter. Researchers identified over 150,000 tweets on the subject of the flooding, which contained information on the location and the extent of damage to the area concerned. The device they used to process the abundance of information was called CarbonScanner, which recognised potential posts on the floods. Over twenty-two thousand photos from the scenes were found from Twitter, Flickr, aerial sources, and the Civil Air Patrol. With this, researchers created an algorithm which would give flood response teams real-time information.
“We looked at a set of images and manually selected areas that we knew had water and areas that had no water,” said Sava. “Then, we fed that information to the algorithm we had developed, and it allowed the computer to ‘learn’ what was and wasn’t water.”
The results of the research have been published in the International Journal of Remote Sensing, concluding that Twitter and other forms of social media could be used to recognise hotspots that satellite imagery alone could not.
Image: Guido Cervone / Penn State
Ed carr
- Edit
It’s great that Twitter and Flickr users can help out in storm situations by doing what they do best – uploading updates and photos. The hard part is that the volume of this ‘big data’ is so overwhelming that some sophisticated algorithms need to be used to sort out real data from the noise, as was done in Colorado, and we need to go from the research lab to practical applications.