Treffer: The Swedish general map as a social document : Exploring change using a conceptual framework of GIScience and a reimagined view of space

The Swedish general map is an important social document with the ability to empower its users through the representation of geographic space it provides. However, the general map produced by Lantmäteriet displays largely the same features as 50 years ago, despite the emergence of new societal challenges over this time. These challenges – such as sustainable urban development and building a safe and resilient society – should ideally be reflected within the general map for it to remain relevant. This study explores how changes can be applied to the general map that ensure it remains a relevant social document, using a theoretical background grounded in GIScience and ideas which challenge traditional perceptions of space. Using this as a foundation, a review is performed of the representation of different vector data within the map: polylines representing the Swedish road network, polygons used to visualise built-up areas (tätort), and points representing the location of shelters (skyddsrum). Methods which share a similar ideology around the structure of space are subsequently used to investigate: (a) whether a space syntax analysis can provide an alternative means of generalising road networks between scales, (b) whether tätort can be classified using power-law distributions and revisualised accordingly at intermediate map scales, and (c) whether a heuristic generalisation process guided by head/tail breaks can be used to include skyddsrum within the map. Results show that generalising road networks using space syntax metrics such as connectivity and local depth offers a viable option to support the existing automated process through the hierarchy of segments it provides. The use of head/tail breaks in visualising tätort by population size is shown to present a more appropriate means of classifying this data and introduces an important thematic element to the general map. A heuristic generalisation of skyddsrum is shown to successfully declutter this dataset for incl