Climate-related challenges for cities and their population are manifold. An accurate representation of the urban built-up environment in weather and climate models can improve forecasting of extreme weather events like heat waves, flooding, and high levels of air pollution. The representation of buildings is further important for the study of pedestrian comfort, building design, and energy usage for urban planning applications. In current models, buildings are represented by surface cover and average height, and they are figuratively ”buried in the ground”, since their effects on the atmosphere are only modelled at ground level. In my PhD I investigated the aerodynamic effects of different building forms (shapes and heights) using large-eddy simulations, and developed a distributed drag parameterization that represents the buildings at their actual height. The new parameterization was tested in a case study of the Greater London region with a high-resolution weather forecasting model from the Met Office. The new parameterization appears to capture more realistic features of the urban boundary layer compared to the standard parameterization, and showed an increased spatial variability across London, which better represents different neighbourhoods such as the dense city centre, high-rise building clusters and residential areas.