Leveraging Twitter Trends for Early Flood Detection: A Case Study of Ruislip, UK

Flooding poses significant risks to communities, necessitating timely and effective warning systems [1]. Social media platforms like Twitter provide a real-time avenue for gathering public insights during such events [2-3]. This study investigates the relationship between flood warning and alert systems announced by the Environment Agency and X(Twitter) trends for the specific area of River Pinn, located in Ruislip, London, UK.  This study employs a systematic approach to explore the interplay between social media activity and flood-related warnings. Keywords such as rainfall, rain, flooding, and flood were identified and used to extract relevant tweets associated with the River Pinn, Ruislip, UK. Data collection involved geotagging techniques and temporal filters to ensure spatial relevance and focus on periods of flood warnings issued by the Environment Agency. The extracted tweets were analysed for temporal trends and spatial distribution to assess their alignment with rainfall events and flooding status.

To investigate the temporal dynamics, cross-correlation analysis was performed between the volume of Twitter activity and the timeline of actual flooding events. Rainfall data from official meteorological sources were also incorporated into the analysis to ensure accurate mapping of precipitation to flooding events. The study further examined whether Twitter activity could act as a predictive tool, evaluating how far in advance users' tweets reflect flood-related concerns compared to observed flood warnings.

The analysis revealed a 30-minute lag between the onset of rainfall and the appearance of related Twitter activity, indicating that social media trends align closely with the progression of real-world weather conditions. More notably, Twitter users exhibited the ability to predict potential flooding events up to one hour in advance. This anticipatory behavior suggests that individuals, through collective observation and situational awareness, recognise the likelihood of flooding before it becomes a reality.The spatial distribution of tweets also highlighted localised concerns, reinforcing the value of geotagged data in enhancing situational awareness for specific areas like Ruislip. These findings underscore the viability of integrating social media insights into flood warning systems, offering a cost-effective and real-time supplement to traditional methods.

This study demonstrates the potential of Twitter as a dynamic tool for flood detection and early warning, with implications for improving emergency response strategies. By harnessing user-generated data, authorities can enhance the effectiveness of flood management systems and better protect at-risk communities.

References

[1] Piadeh, F., Behzadian, K. and Alani, A.M. (2022). A critical review of real-time modelling of flood forecasting in urban drainage systems. Journal of Hydrology, 607, p.127476.

[2] Piadeh, F., Ahmadi, M., Behzadian, K. (2020). A Novel Planning Policy Framework for the Recognition of Responsible Stakeholders in the of Industrial Wastewater Reuse Projects. Journal of Water Policy, 24 (9), pp. 1541–1558.

[3] Bakhtiari, V., Piadeh, F., Chen, A., Behzadian, K. (2024). Stakeholder Analysis in the Application of Cutting-Edge Digital Visualisation Technologies for Urban Flood Risk Management: A Critical Review. Expert Systems with Applications, p.121426.