EGU23-16719
https://doi.org/10.5194/egusphere-egu23-16719
EGU General Assembly 2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

MICA: A fast and flexible solution for real-time flood mapping – An application on 2015 Cannes Antibes flood

Alexandre Bredimas and Tristan Cambonie
Alexandre Bredimas and Tristan Cambonie
  • Bluemapping, France (abredimas@bluemapping.com)

Fast-flood prediction challenges both scientists and stakeholders. Flood evolution is extremely sensitive to input data regarding rain forecasts and the actual status of infrastructures and soil. Accurate and quick modelling is critical to the responsiveness and decision-making of all stakeholders for an optimal allocation of the resources required to limit the damages to the infrastructure and the risk to the population.

BlueMapping develop a decision-support tool called MICA. MICA is a cellular automata model building on previous academic models, especially CADDIES, with tailored adaptations.

The code of MICA has been industrialised into a high-performance calculation algorithm based on parallel computing using GPUs. It has been deployed on the Amazon Web Services cloud. It provides an efficient, scalable and flexible solution for pluvial flood prediction.

MICA's potential will be illustrated with the test case of fast-flood inundations that hit the watershed of Cannes and Antibes (South of France) in October 2015. The code runs in a few minutes on this 149 km2 watershed. The result will be benchmarked with a standard model and the actual maximum depth measurements.

How to cite: Bredimas, A. and Cambonie, T.: MICA: A fast and flexible solution for real-time flood mapping – An application on 2015 Cannes Antibes flood, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-16719, https://doi.org/10.5194/egusphere-egu23-16719, 2023.