Is Spatial Resolution Critical in Urbanization Velocity Analysis? Investigations in the Pearl River Delta

Grid-based urbanization velocity analysis of remote sensing imagery is used to measure urban growth rates. However, it remains unclear how critical the spatial resolution of the imagery is to such grid-based approaches. This research therefore investigated how urbanization velocity estimates respond to different spatial resolutions, as determined by the grid sizes used. Landsat satellite images of the Pearl River Delta (PRD) in China from the years 2000, 2005, 2010 and 2015 were hierarchically aggregated using different grid sizes. Statistical analyses of urbanization velocity derived using different spatial resolutions (or grid sizes) were used to investigate the relationships between socio-economic indicators and the velocity of urbanization for 27 large cities in PRD. The results revealed that those cities with above-average urbanization velocities remain unaffected by the spatial resolution (or grid-size), and the relationships between urbanization velocities and socio-economic indicators are independent of spatial resolution (or grid sizes) used. Moreover, empirical variogram models, the local variance model, and the geographical variance model all indicated that coarse resolution version (480 m) of Landsat images based on aggregated pixel yielded more appropriate results than the original fine resolution version (30 m), when identifying the characteristics of spatial autocorrelation and spatial structure variability of urbanization patterns and processes. The results conclude that the most appropriate spatial resolution for investigations into urbanization velocities is not always the highest resolution. The resulting patterns of urbanization velocities at different spatial resolutions can be used as a basis for studying the spatial heterogeneity of other datasets with variable spatial resolutions, especially for evaluating the capability of a multi-resolution dataset in reflecting spatial structure and spatial autocorrelation features in an urban environment.