Lecture Series “Robust AI”: Nicole Ludwig

The next lecture in our lecture series will be given by Dr. Nicole Ludwig on the topic of “Assessing Climate Change Impacts: An ML Approach to Multi-Decadal Energy Forecasting”.

When & where:

Thursday, June 20 2024, 4:15pm at Bielefeld University (Room CITEC-1.204) or online. Join us via Zoom https://uni-bielefeld.zoom-x.de/j/64775735478?pwd=TFpEUVFPME5EQXFKMHZHY1ZsM2Y4Zz09


Renewable energy is crucial for a more sustainable future. However, as renewable energy sources depend on the weather, they not only help mitigate climate change but are also directly impacted by it. Climate projections provide valuable insight into long-term weather conditions. These global models provide information on different temporal and spatial scales and are computationally expensive with increasing resolution.
Their spatial data resolution is typically insufficient to estimate wind power potential precisely. Deep-learning-based methods could bridge this scale gap, with models for the subtask of image super-resolution, in particular, showing promise of being transferable to increase the spatial resolution of climate models. To assess future energy, we aim for accurate but also data- and resource-efficient models. We want to use as little data as possible, thus the lowest temporal and spatial resolution, which still gives us the best approximation.
In this talk, we look at which temporal and spatial resolutions are needed to assess future wind power and how different ML models compare on super-resolution and downscaling tasks for future wind power


Nicole Ludwig is leading the independent research group “Machine Learning in Sustainable Energy Systems” within the Cluster of Excellence – Machine Learning for Science at the University of Tübingen. Before coming to Tübingen, she did her PhD in Computer Science in Karlsruhe and Oxford and studied Economics and Computer Science in Freiburg and Oslo.

Nicoles research focuses on machine learning for problems related to sustainable energy systems, especially uncertainty quantification and the relationship between weather, climate and the energy system.