One of the major problems encountered in urban mass transit ridership forecasting is the incompatibility between traffic analysis zone (taz) boundaries and service areas for mass transit routes. This incompatibility results in limited sensitivity of many taz systems to passenger access to transit and often results in erroneous ridership estimates. A method to modify existing taz systems to account forpassenger access to specific transit routes in systemwide mass transit ridership forecasting is documented. Three problems are identified and addressed in the application of this method to conventional forecasting methods that use the four-step modeling process of trip generation, trip distribution, mode split, and trip assignment. The first problem concerns how the zone system can be modified to accountfor passenger access but retain both the basic structure of the tazsystem and a taz-based trip table through the entire four-step process. The second problem concerns how model data may be efficiently and accurately reformatted in the framework of the modified zone system. The third problem concerns the adaptation of the four-step urbantransportation modeling process for execution in the modified zone system--particularly in the stages of mode split and trip assignment. This paper appears in transportation research record no. 1285, Transportation forecasting 1990.
Abstract