ABSTRACT
OpenStreetMap (OSM) has evolved as a popular dataset for global urban analyses, such as assessing progress towards the Sustainable Development Goals. However, many analyses do not account for the uneven spatial coverage of existing data. We employ a machine-learning model to infer the completeness of OSM building stock data for 13,189 urban agglomerations worldwide. For 1,848 urban centres (16% of the urban population), OSM building footprint data exceeds 80% completeness, but completeness remains lower than 20% for 9,163 cities (48% of the urban population). Although OSM data inequalities have recently receded, partially as a result of humanitarian mapping efforts, a complex unequal pattern of spatial biases remains, which vary across various human development index groups, population sizes and geographic regions. Based on these results, we provide recommendations for data producers and urban analysts to manage the uneven coverage of OSM data, as well as a framework to support the assessment of completeness biases.
Subject(s)
Machine Learning , Sustainable Development , Humans , Cities , Urban Population , Spatio-Temporal Analysis , ChinaABSTRACT
In the past 10 years, the collaborative maps of OpenStreetMap (OSM) have been used to support humanitarian efforts around the world as well as to fill important data gaps for implementing major development frameworks such as the Sustainable Development Goals. This paper provides a comprehensive assessment of the evolution of humanitarian mapping within the OSM community, seeking to understand the spatial and temporal footprint of these large-scale mapping efforts. The spatio-temporal statistical analysis of OSM's full history since 2008 showed that humanitarian mapping efforts added 60.5 million buildings and 4.5 million roads to the map. Overall, mapping in OSM was strongly biased towards regions with very high Human Development Index. However, humanitarian mapping efforts had a different footprint, predominantly focused on regions with medium and low human development. Despite these efforts, regions with low and medium human development only accounted for 28% of the buildings and 16% of the roads mapped in OSM although they were home to 46% of the global population. Our results highlight the formidable impact of humanitarian mapping efforts such as post-disaster mapping campaigns to improve the spatial coverage of existing open geographic data and maps, but they also reveal the need to address the remaining stark data inequalities, which vary significantly across countries. We conclude with three recommendations directed at the humanitarian mapping community: (1) Improve methods to monitor mapping activity and identify where mapping is needed. (2) Rethink the design of projects which include humanitarian data generation to avoid non-sustainable outcomes. (3) Remove structural barriers to empower local communities and develop capacity.
