Distance Matrix for Aarhus Civil Defence Shelters and Buildings, 1985

The Distance Matrix Aarhus (DMA) dataset is a relational table of linked building–shelter pairs with associated Euclidean and network-based distance measures, created by the MELICA Project (Modelling Everyday Life in Cold War Aarhus) at Aarhus University, 2024–2027. It was produced specifically to support place-based spatial equity analysis of the Cold War civil defence shelter network in Aarhus, Denmark, as of 1985. The dataset was generated from three source layers: (1) georeferenced civil defence shelter locations — public Betondækningsgrave (BDG), combined-public shelters Kombinerede offentlige beskyttelsesrum (KOB), and private Sikringsrum (SR) — digitized from the Aarhus Civil Defence Commission (CDC) collection held at the Aarhus City Archive; (2) building centroids for structures predating 1986, derived from the Danish Building and Dwelling Register (BBR/DBDR); and (3) a 2025 road network clipped from OpenStreetMap (OSM) to the 1988 urban boundary of Aarhus, digitized from a historical map of the city and included here. Distance calculations — both Euclidean straight-line and routed network distances — were performed in R using the sfnetworks, RANN, and data.table packages. The scripts used to generate the dataset are available via GitHub: https://github.com/MELICA-AU/spatial-equity/blob/master/02_DistanceMatrix.Rmd. In its current version, the dataset consists of one row per building–shelter pair, covering 1,698 shelters (331 BDG, 34 KOB, 1,333 SR) and all 40,588 buildings within the 1988 Aarhus boundary with a construction date prior to 1986. Each record links a building to a shelter and records both Euclidean and network distance, as well as the shelter's rank by proximity for that building. Format The dataset is published as a single CSV file (dma1985.csv), loadable directly into R or Python. In Python: dma = pd.read_csv("https://zenodo.org/record/[RECORD_ID]/files/dma1985.csv?download=1"). Column definitions are provided below and in the accompanying metadata file. Column descriptions building_key — integer key identifying the building in lieu of the alpha-numeric building_id shelter_key — integer key identifying the shelter provided in lieu of the occasionally alpha-numeric shelter_id net_m — routed network distance in metres between building centroid and shelter centroid, calculated via OpenStreetMap road network euclid_distance — Euclidean (straight-line) distance in metres between building centroid and shelter centroid shelter_rank — rank of the shelter for this building by Euclidean distance (1 = nearest) shelter_net_rank — rank of the shelter for this building by network distance (1 = nearest)cal building_id — UUID of the building record in the Danish Building Register (BBR) (links to building register) shelter_id — integer identifier of the shelter in the MELICA shelter dataset (links to shelter register) shelter_capacity — number of persons the shelter is rated to accommodate shelter_type — type of shelter: BDG (Betondækningsgrave, public), KOB (Kombinerede offentlige beskyttelsesrum, combined-public), or SR (Sikringsrum, private) Further reading on the dataset creation and methodology: Sobotkova, A, Rosanna Farboel, and Aiswarya Roy (under review), "Shelter for All? Spatial Equity and Civil Defence Planning in Postwar Aarhus." Journal of Digital History (forthcoming). [link to paper when published] Notes on distance measures Both Euclidean and network distances are provided because they bracket the plausible range of historical pedestrian travel. Euclidean distances represent an optimistic, obstacle-free scenario; network distances approximate actual walking routes but may overestimate friction in the historical city centre, where pedestrian shortcuts not captured by OSM were common. The 1988 urban boundary was digitized from a historical map of Aarhus, however, this urban boundary was not necessarily used by civil defense during their planning.. No digital road network from the 1980s exists; the modern OSM network is used as a conservative spatial proxy, justified by the enduring structure of Aarhus's core street grid. Disclaimer The shelter location data was digitized from archival sources by the MELICA project team and cross-verified against historical maps, archival photographs, and satellite imagery. Building data originates from the Danish Building Register (BBR), provided by third parties. The MELICA project does not vouch for the completeness of the shelter or building records. Users are encouraged to consult the primary archival sources at the Aarhus City Archive and the project GitHub repository for full methodological details.