A Novel “Ghost City” Phenomenon Identification Approach Based on Multi-source and Multitemporal Remote Sensing Data

Abstract

“Ghost city” phenomenon refers to a phenomenon of serious imbalance between land urbanization and population urbanization in China. It is necessary to identify and evaluate the change of this phenomenon in long time series, so as to prevent and control its tendency. In this paper, we proposed a novel ghost city index (NGCI) that combines the change of nighttime light intensity in the built-up area featuring spatial-temporal characteristics, the change of the plaque area in the built-up area, and built-up area-population proportion with present situation. These parameters considering the spatio-temporal variations from multi-source and multi-temporal remote sensing data are introduced in the NGCI in comparison to the conventional GCI based on statistical yearbook data. Moreover, the emergence and development processes of “ghost city” phenomenon are taken into account in long temporal series to realize the fusion of statistical analysis and the remote sensing technology as well as the present situation and the spatio-temporal characteristics. In this paper, the rationality of identifying NGCI has been conducted with the development land per capita in all cities subordinated to Guangxi Zhuang Autonomous Region in 2012 and 2015 as the reference standard. Comparing to the conventional GCI, NGCI identifying results were highly in consistent with the numerical distribution and variation trend of development land index per capita towards the identification of “ghost city” phenonmenon.