A Planning based Evaluation of Spatial Data Quality of OpenStreetMap Building Footprints in Canada

Abstract

OpenStreetMap (OSM) is an editable world map where users can create and retrieve data. Building footprints are an OSM dataset that is of particular interest, as this data has many useful applications for planners and academic professionals. Measuring the spatial data quality of OSM building footprints remains a challenge as there are numerous quality measures that can be used and existing studies have focused on other OSM datasets or rather a single quality measure. The study performed in this thesis developed a set of ArcGIS models to test numerous spatial data quality measures for OSM building footprints in a sample of mid-sized Canadian municipalities and gain a comprehensive understanding of spatial data quality. The models performed tests by comparing to municipal datasets as well as determining other quality measures without a reference dataset. The results of this study found that the overall spatial data quality of OSM building footprints varies across mid-sized municipalities in Canada. There is no link between a municipality’s location or perceived importance and the level of spatial data quality. The study also found that commercial areas have a higher level of completeness than residential areas. While the models worked well to test numerous spatial data quality measures for building footprints and can be used by others on other building footprint datasets, there exist some limitations. Certain tests that identify potential building footprint errors need to be checked to see if they are indeed errors. Also, the models were not able to measure any aspect of shape metrics. Suggestions for further studies include measuring shape metrics of building footprints from OSM as well as encouraging and subsequently monitoring OSM contributions in a particular area.