Abstract:
In recent years, there has been a significant amount of interest in accounting for differences between peak and off-peak periods in travel demand models. This interest has stemmed from many factors, including the need to consider time of day effects in analyses of congestion, air quality, transportation demand management, and other issues that are among the main focuses of transportation planning today. Many urban areas have introduced time of day components into their travel model systems. The ways of performing time of day modeling vary in many ways, including the point in the modeling process where daily trips are split into peak and off-peak periods, the number and definitions of time periods into which the average daily analysis period is divided, and the handling of peak spreading. This paper discusses the similarities and differences among some of the time of day modeling procedures that have been recently implemented in U.S. urban areas. Included is a comparison among urban areas of the percentages of trips by purpose and direction for morning and afternoon peak periods. Also discussed are many of the key issues facing modelers incorporating time of day into their processes and how these issues are being handled. The issues include: how feedback of travel times from trip assignment into earlier model steps is handled; whether daily trips are converted to peak and off-peak trips before or after trip distribution and mode choice; how the use of regional peaking factors affects localized peaking phenomena; the spreading of travel demand from the peak hour to the rest of the peak period; and the spreading of travel demand out of peak periods into off-peak periods. The advantages and disadvantages of the alternative ways in which time of day models address these issues are presented.
