Estimation of sight distance on highways with overhanging elements

Sight distance is a key factor in road safety. Sight distance estimation is usually performed on a digital terrain model (DTM). A DTM is a 3D representation of the terrain surface which depicts exclusively t he elevation of the bare ground. However, the reality contains many more elements influencing sight distance than the bare ground. Features such as vegetation, traffic signs, buildings and many other elements are not included in DTMs. The first approximati on solution involves the use of digital surface models (DSM), which comprise these roadside features. The use of geographic information systems (GIS) in highway safety research is interesting due to the easy integration of different factors (accidents, tra ffic data, road features) on the same software platform. However, sight distance studies based on DSM may lead to biased results when there are elements (for example, overpasses, tre e branches or cantilever signals) overhanging the roadway that cannot be interpreted by those models. This paper proposes a method to overcome such difficulties. The proposed solution is based on GIS tools (lines of sight) and multipatch datasets, which repre sent roadside features adequately, added on the DTM. It has been applied in a rural highway located in the Region of Madrid (Spain)