New paper on mapping 3D building structure at a continental scale
Using a huge number of satellite imageries, IVM researchers Mengmeng Li, Elco Koks and Jasper van Vliet are now working collaboratively towards a more nuanced representation of human settlement at very large scales, and they just hit a milestone.
05/14/2020 | 3:39 PM
A recent study led by Environmental Geography group is published in Remote Sensing of Environment, entitled ‘Continental-scale mapping and analysis of 3D building structure’.
This paper goes beyond the mainstream of merely mapping urban extent in the remote sensing community, and as such, it facilitates a more comprehensive analysis of urban environment at a very large scale. Specifically, random forest models are used to predict the 3D building structure for Europe, the US, and China, based on multisource data such as optical Landsat and Sentinel SAR. In particular, building footprint, height, and volume are predicted separately for the three regions combined, and all the R2 values are larger than 0.81.
Further analysis shows that buildings in China tend to be taller on average (10.35 m) compared to Europe (7.37 m) and the US (6.69 m). At the same time, the building volume per capita in China is lowest, with 302.3 m3 per capita, while Europe and the US show estimates of 404.6 m3 and 565.4 m3, respectively. The continental-scale 3D building structure data can be reused in many applications, such as spatial planning, urban thermal analysis, and land use projections.
These data are freely available upon application (https://cscproject.github.io).