Using Social Media Platforms to Model Wildfires

What is it about?

During wildfires, individuals post a subjective account of their surroundings on social media platforms. These individuals can be treated as sensors, responding to physical stimuli and returning a text reading containing local information about the event. Information can be extracted from these posts/readings to model these events from a social perspective. This creates a network of human sensors on the social media platform. Networks on different platforms can be combined to overcome the demographic limitations and biases caused by using a single platform.

Featured Image

Photo by Mike Newbry on Unsplash

Photo by Mike Newbry on Unsplash

Why is it important?

Currently, wildfire modelling mostly involves large, computationally heavy simulations and relies on satellite data. Satellite data is sometimes not immediately available, and the large simulations take a long time to run and sometimes do not capture the extreme behaviour seen in wildfires. They also largely do not account for social factors such as human mobility issues and danger to the public. By using social networks, we can stream information in real-time, and immediately analyse this highly social data, in order to make predictions on these events and ultimately identify areas of danger & interest to emergency services.

Perspectives