A crucial step for the safety verification and validation process of Highly Automated Vehicles (HAV) is the definition of the Operational Design Domain (ODD). The ODD specifies the space of operation for the system under consideration, mainly based on the traffic environment and the subject vehicle behavior. Current approaches are able to describe the static traffic environment by representing the individual elements of the scenery, but they do not manage to identify the behavior-relevant information from these elements. This information is important for the functional specification of the system. We introduce a structure of behavioral attributes that assigns the behavior relevant information to the scenery. This enables the capturing of the limits of the behavioral demand of the scenery, which serves as a basis for the Behavior-Semantic Scenery Description (BSSD). The advantage of this behavioral abstraction process is a reduction in complexity while preserving the behavior-relevant information content. We illustrate this advantage by applying the behavioral attributes to different real-world intersection examples.