Jonathan Critchley

Users of social media such as Twitter are generating vast amounts of geographically referenced information. The widespread availability of such passively contributed geospatial information could benefit planning and decision-making by governments and the public.
We hypothesized that Twitter could act as a source of valuable geospatial environmental observations in order to prepare for, and respond to, exceptional events. We test this hypothesis in the context of extreme heat events in Toronto. Heat-related tweets were captured using Twitter’s Streaming API and grouped by 1-hour intervals. The relationships between heat-related tweets occurring during activity spikes and time of day, temperature, and total Twitter activity were assessed. In addition, the Twitter Streaming API was integrated with an argumentation mapping tool illustrating opportunities arising from location-based social media services to deliver timely and free geographic information.
The analysis determined that passively contributed data were not suitable for professional applications based on the context of the case study. It was concluded that actively volunteered geographic information is currently required for planning and decision-making, and that social media services provide suitable infrastructure for the production and delivery of such information.