A Spatial Latent Profile Analysis to Classify Land Uses for Population Synthesis Methods in Travel Demand Forecasting

In this paper, a new land use classification method is explored for its utility in explaining travel behavior and as a new dimension in population synthesis for travel demand forecasting. This method is based on latent profile analysis applied to 17 business establishment indicators for each of the more than 20,000 block groups in California. The method reproduces the four types of land use environments (urban, suburban, exurban, and rural) identified in a previous paper, and improves our ability to create a finer-grain geographic classification based on land use. It also offers similar indications about the difference between urban dwellers (that make more trips but travel shorter distances) and rural residents (that make fewer trips but with more vehicle miles traveled).