Developing countries are undergoing fast motorization accompanying with rapid economic growth and urbanization. Many cities face increasing traffic congestion and air pollution. Understanding travel behavior is critical to improve urban transportation, environment and public health. This research takes Chengdu, a typical Chinese mega-city with traffic congestion and air pollution, as a case study to discover the relationship among land use attributes, socio-economic characteristics, car ownerships and travel behavior.

The dataset used in the research is travel behavior and attitude survey collected in 2005. It does not have geocoded network and land use that provide a skeletal structure to estimate demand models. This research endeavors to use the available resources and tools to complete the missing level of service information that is the basic input for demand models. The objective of the research is to develop joint choice models with measurement equations that account for measurement errors and behavioral process, and thus provide input to planning and policy analysis.