Development of a Feeder Route Cyclist Demand Model for Dense Urban Areas

Abstract

This thesis provides a new way of forecasting bicycle demand in the urban core by using a feeder system approach. It uses a technique of predicting demand at a location for all of the demand feeding into that point. No other published research has been found to use bicycle count data flowing from upstream to downstream locations in this manner to forecast demand by using the socioeconomic and land use characteristics associated with these locations. This study, in particular, examines the behavior of bicyclists in the District of Columbia. This thesis forecasts a statistical relationship between socioeconomic characteristics and land use characteristics by using GIS and regression modeling techniques in order to determine what factors influence cycling behavior. Previous bicycle demand models fail to provide an accurate method of forecasting the demand because they do not consider the socioeconomic and land use characteristics associated with the flow of cyclists on particular bicycle lanes. This thesis demonstrates the development of two models which both prove to provide better results than the traditional “Point Location” Model (which is also modeled in this thesis for comparison against the two new methods). City officials, planners, and engineers should use the feeder system model so that they can make accurate predictions and design and construct bicycle facilities for the future urban area.