For the design of future climate resilient cities, long-term climate scenarios at high resolution over cities are needed. Current regional climate models, however, lack the representation of urban areas due to their coarse resolution and unsuitable physical parameterizations for the simulation of urban processes. The aim of this fellowship is to acquire knowledge on the impact of climate change on cities by implementing an urban signature into climate data using a novel state-of-the-art method. Different urban vegetation scenarios will be applied to investigate whether vegetation is an effective measure to reduce urban warming in the long term. First, novel climate data with urban signature at sub-urban scale (1 km) for 20 European cities will be created and compared to the current climate data without the urban signature. Secondly, urban greening scenarios will be modelled to investigate the impact of vegetation in the 20 cities under 1.5 °C, 2 °C and 3 °C global warming. Lastly, a novel machine learning algorithm, which will be computationally less expensive than the current technique, will be developed and tested to obtain climate projections at 1 km horizontal resolution that cover all European cities. This work will be accomplished with the support of experts in complementary fields (machine learning, regional and urban climate modelling). Finally, this work will result in a novel approach to create climate data and state-of-the-art climate projections over cities.