Examining income segregation within activity spaces under natural disaster using dynamic mobility network

Rising income segregation perpetuates socioeconomic inequalities, preventing us from developing more sustainable environments. In the literature, little is known about the spatiotemporal dynamics of income segregation under natural disaster. To address this gap, this paper aims to quantify how people experience income segregation in the context of a disaster through their mobility patterns. Specifically, we propose a generalized framework to construct a dynamic mobility network that describes the footprint of residents within their activity spaces, leveraging large-scale mobile phone data. The construction of this network is formulated as a nonlinear programming model and solved by the Lagrangian relaxation and numerical iterative method. A case study is presented to examine income segregation of residents in Harris County (Texas) in access to various critical facilities before, during, and after the winter storm Uri. Our results indicate that income segregation behaviors vary at different critical facilities, as facilities with smaller catchment areas are found to have higher degrees of segregation. We also highlight the disparities in access to critical facilities between low-income and high-income neighborhoods. This paper provides a new venue to better understand income segregation behavior under natural disaster through human mobility networks, which could inform equitable resource allocation in disaster management.