Luis Cea, Jerónimo Puertas and Esteban Sañudo from the GEAMA research group at the UDC are leading the project ‘AI4FLOOD – Improving flood forecasts generated with physical-based models using artificial intelligence techniques’ (PID2023-148074OB-I00) funded by MICIU/AEI/and FEDER (EU).
The aim of the project is to improve physical prediction models by combining traditional models with new artificial intelligence (AI) tools to anticipate flooding and minimise its consequences.
For years, the group has been developing software capable of reproducing the physical processes of a river or city during an episode of heavy rainfall, and now AI complements the model by reproducing the behaviour of physical models more quickly. In addition, it can make comparisons with observed results, detect biases and indicate where modelling can be improved to eliminate these errors.
