A realistic and multilevel measurement of citywide spatial patterns of economic segregation based on human activities

Abstract

Research on the realistic and comprehensive identification of citywide spatial patterns of economic segregation is valuable for the sustainable development of cities. The consideration of human activities in segregation research inspires us to develop an alternative method to contribute to this type of research. In our method, we emphasize the combination of collective activity spaces (CASs) and spatial economic data, both of which are obtained from dynamic human activities. We first reveal the realistic use of urban spaces from human mobility patterns to generate multilevel CASs as basic analytical units. Then, we use a type of realistic economic data generated from human activities to measure the segregation level of each CAS. We realize this measurement by tailoring a segregation index, named the Term Frequency-Inverse Document Frequency-Index of Concentration at the Extremes-based (TFIDF-ICE-based) segregation index, for our economic data. Through these methods, we can uncover citywide multilevel spatial patterns of economic segregation realistically and comprehensively. Using Beijing and Wuhan as cases, we demonstrate and discuss the applicability and value of our method.