KIWA - Artificial Intelligence for Flood Warning
Subproject: Artificial intelligence for runoff forecasting and river monitoring
The management of natural disasters, especially heavy rainfall and resulting flooding, places special demands on the emergency services. Therefore, sufficient lead time is desirable in order to get "in front of the situation" and to be able to evaluate options for flood management actions. Furthermore, river observations are required to obtain the most accurate overview of the situation in the event of flooding. The goal of the project is to develop and demonstrate AI-based tools for flood warning and observation to assist response teams in managing large-scale flood disasters.
These tools include data-driven rainfall-runoff models that emulate watershed behavior and can quickly and robustly process rainfall from different data sources (ensemble forecasts, short-range forecasts, observations) for flood forecasting and warning.
Furthermore, AI-based algorithms will be developed for image-based robust quantification of water levels, flow velocities, and discharge from camera images for streamflow observation, allowing for unsupervised and mobile camera operation. The AIs will be integrated into demonstrator for operational flood management. The AI-based demonstrator will be developed and tested in closed collaboration with associated partners from the field.
Link to the project homepage:
http://kiwa.hydro.tu-dresden.de/
| Duration |
April 2021 - March 2024 |
| Funding by | Federal Ministry of Education and Research (BMBF) |
|
Network partner |
|
| Associated partner |
|
|
Funding program |
Research for civil security 2018 to 2023 |
|
Funding measure |
Artificial intelligence in civil security research |
|
Funding code |
13N15542 |
|
Additional information |
https://www.sifo.de/en/index.html |
| Staff |
|
