Abstract
Estimation of the built-up areas is fundamental to studying urbanization and the concomitant impacts on our rapidly changing planet. Local-scale mapping of the built-up areas elucidates spatially distinct patterns of urban densification as well as peri-urban growth, where the most socially vulnerable population traditionally resides, and helps to ensure sustainable and equitable urban development. Due to the scale and spatial dependency of urban processes, most land use and land cover (LULC) data produced at national and regional scales cannot adequately capture this local variation which is readily observed in very high-resolution (≤0.5 m) remotely sensed images.
Our study investigates whether human settlement data derived from very high-resolution images provide unique understanding in the mapping of built-up areas and further the knowledge of human signatures at local levels. We selected two disparate geographies, Egypt and Taiwan, for which we analyzed four datasets representing human settlements at different spatial resolutions. Our analysis of urban morphology is based on aggregation, complexity, and contiguity of built-up areas on these settlement data and conducted at multiple spatial scales corresponding to the original resolution of the datasets. The results indicate that estimates of the total built-up area are severely misconceived, with most anomalies occurring along fringe areas. This work also illustrates the potential of high-resolution datasets to provide new insight into urban dynamics, through determining new measures of built-up area and identifying complex urban and peri-urban patterns that were previously undetected.
