Annually modelling built-settlements between remotely-sensed observations using relative changes in subnational populations and lights at night

Measuring the contribution of built-settlement data to global population mapping

“…Human settlement and population dynamics are more important than ever to understand ( Ehrlich, Balk, & Sliuzas, 2020 ; Zhu et al., 2019 ) as an additional 13 percent of the world’s population will live in urbanized areas by 2050, with most of this growth occurring in low-to middle-income countries ( Angel, Parent, Civco, Blei, & Potere, 2011 ; United Nations, 2018 ). Most of this projected growth will not occur in the largest cities, but rather it will occur in small to medium sized settlements ( Cohen, 2004 ), which are typically underrepresented in various measures and counts including censuses ( Leyk et al., 2019 ; Tatem, Noor, von Hagen, Di Gregorio, & Hay, 2007 ) and remote-sensing (RS)-derived representations of settlements ( Kuffer, Barros, & Sliuzas, 2014 ; Kuffer, Pfeffer, & Sliuzas, 2016 ; Nieves et al., 2020 ; Pesaresi et al., 2013 ; Weber et al., 2018 ). This projected growth has implications for sustainable development ( Ehrlich et al., 2020 ), which has been noted in the 2030 Sustainable Development Goals (SDGs) ( United Nations, 2016 ).…”

“…However, since 2010, a new class of globally available and consistent RS-derived representations of built-settlement, have become available at single and multiple time points ( Corbane et al., 2017a ; Esch et al., 2013 ; Esch et al., 2018a ; Facebook Connectivity Lab, 2016 ; Microsoft.BuildingFoot, 2018 ; Pesaresi et al., 2013 ; Pesaresi et al., 2016 ). Built-settlement (BS) is defined as above ground structures that can support human habitation and related economic phenomena ( Florczyk et al., 2019 ; Nieves et al., 2020 ; Pesaresi et al., 2013 ). The concept of BS addresses the “distribution of buildings by which people attach themselves to the land” ( Ehrlich et al., 2020 ; Stone, 1965 ) and these data are better able to differentiate between buildings and other aspects of the built environment, such as road ways or parking lots.…”

“…The aforementioned settlement and population modelling needs, combined with the current temporal limits of BS datasets, have prompted some to interpolate BS extents in a globally consistent manner ( Nieves et al., 2020 ). These efforts produce annually estimated BS extents while expanding the temporal frequency and coverage of these BS datasets while maintaining its dataset specific definition of BS ( Nieves et al., 2020 ). A larger question, accompanying any further application of modelled data, is how these interpolated settlement extents contribute to subsequent modelling applications when lacking comparison to a ground truth.…”

High-Resolution Gridded Population Datasets: Exploring the Capabilities of the World Settlement Footprint 2019 Imperviousness Layer for the African Continent

“…For the most part, the recent emergence of global (or near-global) built-area datasets that accurately describe the extent, location, and characteristics of human settlements has been exploited in the production of new population grids, resulting in improved quality, accuracy and spatial resolution. Representative examples include recent population distribution datasets that have been produced on the basis of the World Settlement Footprint 2015 products (WSF2015 and WSF2015-Density) [32]; the new WorldPop Sub-Saharan gridded building datasets [33][34][35]; or through the joint analysis of high-resolution binary built-area products [36,37], such as the Global Urban Footprint [38,39], the High Resolution Settlement Layer [23,40] and the Global Human Settlement Layer [41,42], respectively. Here, the particular focus placed on built-area datasets for population modelling arises from the fact that different research has demonstrated that when built-area datasets are used to restrict the distribution of the population, the final products deliver better qualitative and quantitative results in comparison to those models where the datasets are not included [37,43].…”

Research on Sustainable Development Ability and Spatial-Temporal Differentiation of Urban Human Settlements in China and Japan Based on SDGs, Taking Dalian and Kobe as Examples

“…e data come from interviews based on the subjective feelings of residents [2], questionnaires [3], and official statistical data [4]. With the rapid development of science and technology, data acquisition methods are more advanced, such as night light data [5], vegetation coverage data [6], land cover data [7], and other remote sensing image data, as well as electronic map points of interest [8], cell phone signaling data [9], social network data [10], real estate network data [11], and other network data, which enrich the new data environment composed of big data and open data. In terms of research methods, with the diversification of research data, research technical means are gradually enriched, mainly including quantitative analysis, such as entropy method [12], analytic hierarchy process [13], Delphi method [14], principal component analysis [15], structural equation model [16], and the combination of GIS spatial analysis and geographic measurement model, which is currently widely used [17,18].…”