High resolution census grids provide useful inputs for transport and urban modelling, but confidentiality measures and heterogeneous tables introduce inconsistencies and missing demographic detail. We develop a framework that reconciles multi scale census grids and enriches them with additional attributes while preserving additivity and privacy constraints. The approach restores consistent totals across resolutions and reconstructs detailed demographic profiles. Applied to Germany’s Census 2022, it yields consistent 10 km, 1 km, and 100 m grids and handles structural gaps explicitly. The resulting datasets offer reproducible, demographically rich inputs for agent based mobility models. The pipeline is available at github.com/TUBS-IVS/cleancensus.
Petré et al. (Thu,) studied this question.