Multitemporal grid based analysis of the Global Human Settlement Layers : SMOD taxonomy of change

Urban transformation is an emerging topic in recent years and its socio-economical and environmental impacts have become increasingly evident to scientists and policymakers. The ability to analyse and understand this process was limited by the lack of comprehensive datasets and formalised methodologies. The Global Human Settlement Layer (GHSL) suite accommodates this need by providing worldwide, multi-epoch information on human settlements. In particular, the GHSL suite includes three main layers that map built-up density (GHS-BUILT), resident population density (GHS-POP) and human settlements (GHS-SMOD). These grid-based layers are produced in a consistent and harmonised way for four epochs (1975-1990-2000-2015) allowing the comparison between different periods. In this technical report, we present a formalised methodology (workflow) to characterise and analyse the dynamics of settlements. This workflow proposes a taxonomy of all possible GHS-SMOD classification combinations between two epochs; it quantifies and analyses changes of built-up surface and population density in each taxonomical class within a selected area. We show the workflow capability by using the region of New York (United States of America) and Colombia as case studies. The presented workflow has a direct application to characterise urban dynamics using the GHSL. Such approach supports the assessment of urbanization processes, by monitoring urban expansion, contraction and rural-urban transitions, and by measuring sustainable urban development metrics. The application of this workflow allows taking into account, both and separately, the net and the gross change in built-up surface and population density within an urban settlement evolution. Moreover, the application of this workflow to the GHSL suite benefits of the full range of GHSL features (global coverage, data consistency and open and free data format).