Potential and limitation of using OSM data for the creation/validation of Land Use/Cover maps

Sunday 16:00, S.1.3

Cidália C. Fonte 30 minutes

University of Coimbra, Department of Mathematics/Institute for Systems Engineering and Computers at Coimbra, Coimbra, Portugal

OpenStreetMap (OSM) provides a large amount of valuable data in many parts of the world, that is foreseen to increase, becoming even more useful, especially for the regions where it is most needed. The richness of OSM data is associated to the hundreds of pairs key/value identified in the OSM Map Features Wiki and the additional ones volunteers may create. The OSM Map Features include the keys "Landuse", "Natural" and "Waterway", which are particularly useful to extract Land Use/Land Cover (LULC) information. However, it is also possible to extract additional and more complete LULC data from many other features, which may be used for 1) the creation of LULC maps [1] ; 2) to validate LULC maps [2]; 3) to train classifiers to classify satellite imagery [3], or 4) to create hybrid products, adding information extracted from OSM to enrich existing LULC maps [4]. However, several problems and limitation for this use of OSM can be identified. Among these are: 1) the existence of key/values not listed in the OSM feature map wiki, raising problems to the development of automated approaches; 2) the difficulties in associating some key/values to only one LULC class; 3) the overlap of features associated to different LULC classes. These difficulties will be highlighted in this presentation, and suggestions of approaches to, if not solve, at least decrease the importance of some of these factors. These include the development of methodologies to extract additional information from the data available, but also the inclusion of possible additional features in the OSM data collecting protocol.

[1] Fonte, C., J. Patriarca, M. Minghini, V. Antoniou, See, L., M.A. Brovelli, 2017, Using OpenStreetMap to create land use and land cover maps: development of an application. In: Volunteered Geographic Information and the Future of Geospatial Data, ed: Campelo C., Bertolotto, M., Corcoran, 113 - 137. Hershey: IGI Global. [2] Fonte, C.; Martinho, N., 2017, Assessing the applicability of OpenStreetMap data to assist the validation of land use/land cover maps. Int J GIS, 31, 12: 1 - 19. [3] Schultz, M.,Vossa, J., Auera, M., Carter, S., Zipf, A., 2017, Open land cover from OpenStreetMap and remote sensing. Int J Appl Earth Obs Geoinformation, 63, 206-213. [4] Fonte, C., Minghini, M., Patriarca, J., Antoniou, V., See, L., Skopeliti, A., 2017, Generating Up-to-Date and Detailed Land Use and Land Cover Maps Using OpenStreetMap and GlobeLand30. ISPRS Int J of Geo-Information, 6, 125.