Exploring Residential Heterogeneity Through Multiscalar Lens: A Case Study Of Hangzhou, China

The pattern, process, and mechanism of residential heterogeneity vary significantly with different geographical scales. However, most traditional methods ignore the checkboard and modifiable areal unit problem (MAUP), which may cover up the complexity and hierarchy of social space. Taking Hangzhou city as an example, a multiscalar method was proposed based on the information entropy theory to estimate residential heterogeneity and its scale sensitivity. Based on the sixth population census of Hangzhou and the housing price database of 6,536 residential districts from 2008 to 2018, we explore the scale effect and dynamic characteristics of residential heterogeneity. The results of spatial simulation and geostatistical analysis based on Python Spatial Analysis Library (PySAL) module show that the multiscalar algorithm better presents the real segregation pattern than traditional method, which is one of the new models and technologies in urban geography complex system. Exploring residential heterogeneity through multiscalar lens provides an important basis for the gradual and refined urban renewal.