Simulated Effects of the Location of High-Speed Rail Stations on Land Development: Case Study of the Lisbon Metropolitan Area, Portugal

Abstract

This paper studies the effects of high-speed rail (HSR) on land cover change in the Lisbon metropolitan area (LMA) in Portugal according to a bilevel cellular automata/agent-based modeling framework. The model incorporates spatial mixed logit models at the local level and panel simultaneous equations models at the regional level. The regional submodel generates socioeconomic activities that are input into the local submodel for land development. The regional activities are also influenced by the local outputs. A backcasting simulation is executed to validate the model on the basis of the actual land cover change from 1991 to 2011. Five scenarios are then designed according to the numbers and the locations of HSR stations in the LMA. The simulation shows that the unbuilt land covers are largely developed by 2031, regardless of the inauguration of HSR. Under the scenarios with HSR, because of the dramatic improvement of accessibility, the land cover change speed is largely accelerated. With an additional station on the southern bank of the Tagus River, the overall accessibility does not further increase from the scenario with only one HSR station. Thus, the total land development remains similar. The enhanced access to the HSR service from the additional station contributes to the redistribution of the land cover change activities by moving them slightly southeastward. The results highlight the effect that station location could have on land cover change, pointing out the importance of carefully choosing the location.