Physical- and Social-Based Rain Gauges—A Case Study on Urban Flood Detection

Abstract

Floods are among the most frequent and costly rainfall-triggered disasters. In this context, geospatial content generated by non-professionals using geolocated systems offers the possibility of monitoring environmental events. This study shows a statistical correlation between in situsensors, radar, Twitter posts, and flooding events. Furthermore, we observed in this study that flooding-related keywords are statistically more significant on flooding days than on non-flooding days and reinforce that Twitter can be employed as a complementary data source for flood management systems.