In recent years, the need to reduce the global warming of the planet has become more imperative than ever. Global warming and, at local scale, Urban Heat Island phenomena are among the primary effects of the increased building carbon emissions. Nevertheless, understanding and controlling the parameters which intensify, or mitigate, the increasing in temperatures in the surroundings of a building are pivotal, as sustainable design can significantly reduce the buildings’ energy demand. Micro-climate simulations can provide more accurate input for building energy simulations since they can accurately simulate the interactions between those parameters to calculate detailed weather data. Despite the increasing knowledge about the significance of the microclimate, energy simulation users still rely on derived, or interpolated weather data from sparsely located weather stations, located generally outside the urban environment. The reason behind this commonly adopted approach is that the generation of microclimate data is costly in terms of time, and currently standards for storing this generated data have not been developed. ENVI-met is a microclimatic simulation software that requires a model of an urban area and weather parameters on its boundaries to generate a large extent of data like air, temperature, relative humidity, wind speed etc. Constructing this model manually contains a number of significant limitations, such as high design cost in time and need for data collection from different sources – thus the chance of design errors is high. In this thesis a novel approach is introduced where the ENVI-met software is used for microclimatic simulations at district scale. However, the input model in this case is created by data extracted from a CityGML-based 3D city model. In addition, the generated microclimatic data is stored back to CityGML, where it can be re-used. The proposed methodology is implemented via a Graphics User Interface, divided in two main phases, serving the required bi-directional data flow. It was designed and implemented based on the following specifications: i) the user involvement in the whole process needs to be minimum, ii) the interface should create simulation-ready input models of various resolutions and iii) it must work with different CityGML datasets.
A data requirement analysis indicated that a CityGML-based city model can feed its data to ENVI-met by the interface, so that the input model required by ENVI-met can be constructed fully automatically. In return, the storage of the generated results data is also possible. Therefore, an automated data flow between a CityGML-based city model and ENVI-met can be achieved, offering the following advantages: i) the ENVI-met input model can be constructed fast and automatically and ii) ENVI-met outputs can be translated to real world coordinates – thus can be visualized and processed in GIS software and ultimately stored back into the CityGML-based 3d city model.

3D computer models are starting to play a more and more important role in our society. Realworldsituations are often too complex to explain in a 2D map and also the interest in virtualreality, serious gaming and other technologies that can be based on 3D computer models, isgrowing.