Abstract: In the process of rapid suburbanization, large-scale affordable housing and new towns have been built in Beijing's suburban areas, which formed a unique suburban living and daily activity space. However, it leads to some urban problems because of the lack of supporting facilities and job opportunities around this kind of suburban mega-communities. For example, the long distance of commuting, transportation congestion and jobs-housing spatial mismatch, which also lower the residents' life quality. With the behavioral turn in human geography, space-time behavior has become one of the important perspectives on studies on urban space. Activity space, which is an important measurement in the study of urban social space, has attracted much attention from scholars at home and abroad. In China, research on activity space has focused its attention on aggregated aspects, using density interpolation based on questionnaire data to measure activity space, which ignores individual differences of residents. Data used in this study are from activity and travel survey based on GPS, which was done in 2010, with two suburban mega-communities of Tiantongyuan and Yizhuang as cases. Methods of case studies, spatial analysis based on GIS and multiple linear regression are used. The study uses standard confidence ellipse to measure individual activity space from a disaggregated perspective. Spatial overlay analysis is used to study residents' activity space of weekdays and weekends, and their use of urban space and community space. The study also analyzes the influencing factors of residents' use of urban space to examine problems emerging in suburbanization of China's large cities.

Key words: suburbanization, activity space, GPS, GIS, Beijing