Numerical models to forecast urban floods

The impact of urban floods on homes, lives and costs is expected to increase with aging drainage systems, climate change and continuous infrastructure construction. Highly accurate numerical models are required to forecast floods and simulate the surface runoff of flood waters over time to provide adequate response mechanisms.

IBM Research is developing computational models using high-order numerical methods to model urban floods. The models use the precipitation data from highly accurate short-term weather forecasting of Deep Thunder and surge data as required from Deep Current. Other features like soil type, surface elevation will also be integrated in these models. The results from these models can potentially help in developing response and resilience strategies to minimize the loss of life and property.


Seshu Tirupathi


Predict the vulnerability of a geographic location to floods in the near future by including:

Example of results

Snapshot of water elevation after a partial dam break. The region marked in red is the dam.

urban flood