Estimation of missing building height in OpenStreetMap data: a French case study using GeoClimate 0.0.1

Abstract. Information describing the elements of urban landscapes is required as input data to study numerous physical processes (e.g., climate, noise, air pollution). However, accessibility and quality heterogeneous across world. As an example, a major open-source geographical project (OpenStreetMap) demonstrates incomplete regarding key properties such building height. The present implements evaluates statistical approach that models missing values height in OpenStreetMap. A random forest method applied estimate based on building’s closest environment. total 62 indicators are calculated with GeoClimate tool used independent variables. training dataset 14 French communes selected, reference provided by BDTopo IGN. An optimized algorithm proposed, outputs compared evaluation dataset. At scale for all cities, at least 50 % buildings have their estimated error less than 4 m (the cities' median heights range from 4.5 18 m). Two (Paris Meudon) demonstrate results deviate main trend due specific fabrics. Putting aside these two communes, when averaged regular grid (100 m×100 m), absolute 1.6 m, 75 cells any city lower 3.2 m. This level magnitude quite reasonable accuracy (at uncertainty equal 5 work offers insights about estimation using methods contributes use datasets software. software produce freely available https://doi.org/10.5281/zenodo.6372337 (Bocher et al., 2021b), can be accessed https://doi.org/10.5281/zenodo.6855063 (Bernard 2021).