Representing Buildings for Visibility Analyses in Urban Spaces

The availability of detailed and precise digital surface models based on LiDaR data allows accurate calculation of visibility analysis even in urban areas. Lately, the viewshed analysis, which is implemented in geographical information systems, is often used to determine the visibility of buildings or other structures in both natural and urban environments. Such utilization of viewshed tool, which is originally designed to assess visibility from point to its neighbourhood, however, brings issues regarding partial visibility of the target that are usually neglected. The core of the problem here is that the target building is often represented as a single point in the viewshed analysis. This simplification can lead to an incorrect assessment of the visibility as the specific point of the building can be invisible for the observer while other parts of the building are visible. To properly analyse visibility of a building it is necessary to consider partial visibility of the target. To allow the assessment of partial visibility more than one point that represents the building needs to be defined. In this contribution, the theoretical aspects of reverse viewshed, an area from which a target point is visible, are considered with a focus on the proper representation of target building in the reverse visibility analysis. A practical study of building visibility is conducted with the building represented as single and multiple points. The results are compared and the differences are explored.