Authors: Yoo Min Park*, Geography & GIS, University of Illinois at Urbana-Champaign, Mei-Po Kwan, Geography & GIS, University of Illinois at Urbana-Champaign
Topics: Spatial Analysis & Modeling, Urban Geography, Geographic Information Science and Systems
Keywords: Multi-contextual segregation; Daily activity; Uncertain geographic context problem (UGCoP); Spatiotemporal approach; Human mobility; Travel diary data
Session Type: Paper
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon C2, Sheraton 3rd Floor
Presentation File: No File Uploaded
This study examines how people experience different levels of segregation in various geographical and temporal contexts in their daily life, using individual daily movement data. While residential segregation has been well studied, little is known about segregation that occurs in non-residential contexts, such as places of work or leisure activities. Emphasizing a need of a comprehensive notion that encompasses the full spectrum of individuals' segregation experiences, this study proposes a new notion of segregation, called multi-contextual segregation. It also presents a new individual-level spatiotemporal measure of segregation that can assess multi-contextual segregation. Using individuals’ activity-travel diary datasets collected in the greater Atlanta region in Georgia, this study finds that people experience varying levels of segregation over the course of a day depending on where they spend time while conducting daily activities. People from different racial groups tend to be more evenly distributed during the daytime (i.e., when the majority of people are away from residential areas) than at night (i.e., in residential areas). However, racial minorities remain highly segregated into the inner city or inner-ring suburbs both during the daytime and at night. In addition, even the same household members experience different levels of segregation throughout a day if they work in different areas. The suggested method addresses several methodological problems in residence-based, aggregate-level measures of segregation (e.g., the uncertain geographic context problem, the modifiable areal unit problem, and the checkerboard problem) and enables a comprehensive understanding of individuals' complex, dynamic segregation experiences.