Pluvial flood hazard assessment from the lens of social media data

Abstract

The rapid nature of urban flooding from intense rainfall means accurate surveying of flood data such as water depths and flood extents are scarce and not readily available, hindering assessment of urban flood risk and validation of models. As the amount of volunteered information created by social media grows, it can be used as an effective tool to track the process of the urban pluvial flood and has great potential to offer localized flooding information to serve modeling and facilitate urban flood hazard assessment. In this study, the utility of social media data in China will be examined based on the 2020 Chengdu storm-flooding and 2021 Zhengzhou severe flooding response activities on Weibo and Tiktok platform. Spatio-temporal patterns of public responses towards urban flooding and impacts were mainly analyzed to help urban flood risk management and a modeling framework through social media data is presented. The temporal evolution of social media activities was investigated to track the flooding process and further compared with observed precipitation data. Spatial information was extracted and the typical hotspots are selected for detailed analysis. Moreover, the modeling framework could compare the simulation results against flooding information identified through spatio-temporal social media data analysis, allowing inundation to be inferred elsewhere in the city with increased detail and accuracy.