Sustainable travel during the Olympic and Paralympic Games : A methodology to model public transport travel for Paris 2024

Abstract: This Master Thesis develops the challenges of travel modeling during the Olympic and Paralympic Games, more specially for the Paris Olympics in 2024. This problem as been set by IDFM (Île-de-France Mobilités), the transport organisation authority of Paris and its region, that has therefore to deal with public travel during the Olympics. A very simplified model was already in use, but is no longer sufficient. The exceptional nature of this event, considered as a mega-event, requires a precise understanding of the subject as well as a different and adaptive modeling process. Thus, this work presents a detailed methodology for public transport travel modeling in Paris and its surroundings during the Olympics. This model will become more and more refined until the end of this mega-event, in order to present results or advert the multiple stakeholders around the topic of the Olympic Games transportation (event organizers, transport operators). The two significant parts of the model are distinguished and described: the Olympic Games related trips and the background demand, which require two very different approaches. The OG demand needs several assumptions which are often in constant evolution: the versatility of the parameters is a very important point to take into account. On the other side, the background demand prediction is a significant challenge because it differs from what is usually done. Both of these parts are adapted from the principle of the four-step transportation model and reuse parts of the IDFM model, ANTONIN 3, specifically calibrated for the Île-de-France region. It is also necessary to conceive with the will to adapt as much as possible the available transport data and the tools already in operation, such as the model already in use. Suggestions for further improvements are also mentioned to refine the results until the final day which will be possible thanks to enhancements of the input assumptions over time, such as ticketing data for instance.