Karolina Pantazatou

Urbanization is driving cities to grow, often through urban densification. Utilising computer-based simulations executed over 3D representations of the city is imperative in addressing the challenges posed by implementing sustainable urban densifications; like ensuring adequate access to daylight and limiting noise pollution. However, since these simulations are usually resource-intensive, timeconsuming, and costly, urban planners are often limited to evaluating only a single densification scenario. Additionally, as consultants usually deliver simulation results in static PDF-reports rather than in machine-readable format, urban planners are prohibited from viewing results from multiple simulations in one single visualization, which is problematic when outputs from different simulations contradict each other for the same densification scenario. Instead, they are limited to reading separate reports and
making decisions based on their expertise and experience, leaving room for human error. Furthermore, even when consultants do deliver simulation output data, they seldomly include enough metadata for the dataset to be interpretable. Finally, most municipalities do not have standard practices allowing the storage of such datasets in locations accessible to all stakeholders involved in the urban planning process, thus further hindering their reuse.
To address these challenges, the aim of this thesis is to enhance the use of solar access and noise simulations in the urban planning process with a particular focus on urban densifications. To achieve this, it investigates ways of reducing the cost of these
simulations by attempting to limit the time required to locate and preprocess suitable input data. Towards that end, this thesis attempts to determine the minimum requirements on semantic 3D city model data that are used as input and make suggestions for updates in measuring guidelines related to national specifications on geodata.
Another way of realising the aforementioned aim is by storing expensive simulation output with descriptive metadata in a standardised way and providing the functionality to link it to the corresponding geometries of a semantic 3D city model. This realisation renders simulation results accessible and reusable while functioning as a platform for supporting the execution of queries combining results from multiple simulations with semantic attributes of semantic 3D city model objects over parts of
or the entire municipality, which is useful for identifying vulnerable areas or areas that should be prioritised in strategic urban planning. This platform does also have the potential to support high-quality 3D visualizations of simulation output as well as visualization tools. Thus, improving communication between stakeholders, promoting data-driven decision making, introducing a higher level of transparency, and promoting citizen participation.