According to a new study, images of floods and other large amounts of water uploaded to the online social image gallery Flickr could help scientists predict floods before they overwhelm flood-prone areas and communities.
Researchers at the University of Warwick studied “tagged” photos on Flickr with words like “river” and “water” which were matched to known flood events, over a period of 10 years.
These photos were then uploaded to a “neural network”; this is a computer model that can be taught to identify patterns in large sets of data.
By studying the 5-day period before and after floods, they noticed that as a flood builds, the water-related tags are used more often with flood-related words like “flood” and “floodplain.”
The researchers then monitor for the same tags but in real time, to determine the likelihood of floods approaching.
This is one of the first examples of using social media to predict flooding rather than reporting on it in real-time.
For the complete article click here.
To visit the Science Magazine website where this article originally appeared, click here.