A sustainable transport policy conjoint with an efficient infrastructure requires sound understanding of future travel demand. One important factor in this respect is the shift in population distribution. State of the art transport models account for changes in a population’s age distribution, yet rarely consider changes in behaviour of individuals and cohorts. It can be shown that this deficit leads to distortions in demand forecasts. To understand changing behaviour, specific data is required. Based on panels and pseudo-panels, changes in different age classes and different segments of travel demand can be analysed, allowing for subsequent identification of growth or stagnation drivers. Moreover, different age groups and life stages can be identified with regard to changes in travel behaviour. These stages can be subjected to further examination. In an ageing society, special consideration is given to the time immediately before and after retirement. Based on a cluster analysis different types of retirees can be distinguished. Associated results show that forecasting travel behaviour with mobility biographies requires complex models as well as detailed panel data. Consequently, age-cohort models were employed in this study owing to their comparably pragmatic nature. With regression analysis, age and cohort effects can be segregated and results can be used for extrapolation of macro figures such as total mileage. Furthermore, they may also serve as input data for micro-simulation models. In this study, a heuristic algorithm allowing depicting activity patterns subject to macroscopic constraints was developed. This methodology is demonstrated through the trip generation of two discrete spatial types: An urban growth area is compared to a rural area with decreasing population. The resulting time variation curves show that cohort effects have a significant impact on future travel demand and should thus be incorporated in transport models. (Author/publisher)
Abstract