In our project we try to predict floods using machine-learning methods and ERA5 reanalysis data. The result will be compared with forecast reruns from the operational Global Flood Awareness System (GloFAS).

We did apply for this challenge because it’s a great opportunity to gain experince in writing project proposals, working in an international collaboration and since we have the needed knowledge in the relevant topics -  machine-learning and atmospheric reanalyses. A further argument was the orientation as open-source development, so that our work is not in vain and as many interested parties as possible profit from it. Our goal until the end of the project, is to create an interesting comparison study between machine-learning methods and to share our findings and experiences in the form of Jupyter-Notebooks.

Further informations:

https://www.ecmwf.int/en/learning/workshops/ecmwf-summer-weather-code-2019