Developing real-time flood monitoring and communication capabilities for cities.


Developing real-time flood monitoring and communication capabilities for cities.


Developing real-time flood monitoring and communication capabilities for cities.

About Us

FloodAware is a multi-university project to assess the effectiveness of several real-time flood detection, reporting, and communication technologies for cities and local communities. The project is supported by the National Science Foundation's Smart and Connected Communities program (award 1831475).





We are testing multiple mounted camera technologies. First, StarDot cameras with RBG, NIR and day/night capability, each augmented with a cellular data modem for continuous accessibility, will be implemented at known flooding locations. They will collect imagery from a targeted field of view at a certain time interval; this time interval may vary depending on actual conditions, e.g, once every 30 minutes during good weather, once every minute during heavy rainfall. The FloodCams will be aimed at a target, such as a storm drain, intersection, or stormwater basin; a white and red striped painted gauge will be placed on the target to provide a positive length scale. We are placing cameras on street light poles in high priority locations where flooding is known to occur, and painting staff gauges on curbs and poles.


We are installing gauge points (striped curbs) in each city, to work with and independently of floodcams. These gauge points will provide targets for (and serve as reminders to) community members to snap and upload flooding images using the mobile app. Additional gauge points are being installed incrementally based on user-reported flooding at un-gauged locations. The gauges can be painted on curbs.


The Mobile Hydrology app will not only provide a crowd-sourced stream of primary flood stage data, imagery, and floodwater presence/absence data, but will also communicate visualizations of rainfall and flood histories to users, encourage users to collect data during heavy precipitation events, and notify users when floods are likely to occur or are currently occurring near their immediate location.


We are integrating information from the CrowdHydrology platform to have the capacity to receive and process flooding data from a variety of existing networks outside of our own. CrowdHydrology is a citizen science-based observation network that collects spatially distributed measurements of stream stage in ungauged watersheds. Users text in water levels at existing gauge plates. The CrowdHydrology integration serves as an important step for integrating important flooding information streams from existing systems. Text messages of water levels are integrated into a publicly available database. The system is actively being used in 8 states and has recorded over 10,000 measurements. It serves as an important information stream where fixed gauges offer sparse coverage.


We are extracting data from social media streams, often called social media mining, which has become an increasingly effective data stream of realtime information. We are extracting two valuable pieces of information from such streams. First, posts to platforms like Twitter and Facebook spike when and where there is something “exciting” to share. Filtering Twitter and Facebook streams by locality, then mining them for keywords and hashtags (e.g. flooding, overflowing, under water) can help highlight areas of particular interest within a local flood stage model. Second, users often post images in addition to their comments which can be extracted and analyzed for water level information.