iTURF

Integrated Twitter Urban Real-time Flood Model (iTURF)
iTURF logo

Project contact Luke Smith

Harnessing crowd-sourced data taken from social media for real-time incident management during flooding.

The iTURF model will use data provided by the public to simulate urban flooding caused by intense rainfall in real-time. The results are then fed back to the public through social media and an online map showing the most likely affected areas.

Possible applications include:

  • Real time incident management, identifying roads and pathways which have flooded automatically 
  • Correlating crowd source data against pre-computed flood risk areas to determine other areas likely to be affected

The iTURF model can be applied to events where data collection is ordinarily difficult such as flooding from intense rainfall (FFIR). These types of events are difficult to forecast, arrive with little warning and pass quickly, making data collection through conventional methods (ie. wrack and water mark surveys) ineffective.

Social media can house a large quantity of relevant information during rainfall and flooding events. During a convective storm event in Newcastle on 28th June 2012, more than 1850 Tweets could be identified on the subject of rainfall intensity and flooding. Many of these contained photographs, spatial locations in the form of named buildings and areas of the city. Sometimes there were also indications of flood depth (eg. 'Knee deep').

The iTURF model will harness this real-time data to simulate urban flooding, and will feed results back to the public.

iTURF: research.ncl.ac.uk/iturf/