KIB-UDZ
The city of Speyer is one of the hottest cities in Germany – in 2021 it had the most hot days in the country. With ADAPT, a tool is being created that solves a previously unsolvable task: proven but computationally intensive climate models normally take a long time to deliver results. ADAPT trains an AI to imitate the behaviour of these models and thus delivers simulations of heat and heavy rainfall considerably faster. Local authorities can therefore play through at the push of a button how, for example, a new green space or additional trees would affect heat stress in a neighbourhood – and can do so without needing expensive data centres or specialist staff of their own. It is based on the City of Speyer’s exceptionally detailed data. The freely accessible, platform-independent data and software infrastructure, which can be integrated into existing urban digital twins, is intended to serve as a model for other local authorities in Rhineland-Palatinate.
Project partners:
Deutscher Wetterdienst (Koordinator), Stadt Speyer, Deutsches Forschungszentrum für KI (DFKI), Pecanode GmbH / PALM4U, Technische Universität München (TUM)