Ground access mode choice models of Istanbul airports

One of the important parameter of airport landside planning is providing access alternatives to air passengers. As a result of the characteristics of the travellers, duration of the travel and differences in the current modes and, the factors in airport travel are importantly different from those in the urban transportation models for the regular travels. Most of the travel is conducted by the visitors to airports; in comparison to other urban travels, passengers prefer to travel in groups. Also, their trips for airport access purposes have lesser frequency than the other urban trips. They carry their luggage with them and they are mostly away for a couple of days or sometimes for weeks. In this study, two airport in İstanbul; Sabiha Gökçen International Airport (SAW), Istanbul Atatürk International Airport(IST) were investigated and access mode choice models were built by using the air passenger survey data. Multinomial Logit Model was used to analyze mode choice. Passenger characteristics and trip characteristics are the major parameters of models. Results of the analysis show that travel time, trip purpose (Business/Nonbusiness), travel cost, flight route (Domestic/International) are the most important parameters of access mode choice. For instance, being international or business passenger for SAW model and number of passenger for IST model decreases probability of choosing public transit relative to auto. Also, distance has a negative effect on choosing taxi relative to auto for both airport. So, if distance increases, proportion of choosing taxi decreases. Furthermore, in this study, two factors which did not used for any models in literature, were used for mode choice models. One of them is seasonal effect (winter/summer) which was used for IST model and winter increases choosing taxi compare to auto. Other one is Weather effect (rainy/not rainy) which was used for SAW model and if weather is rainy choosing public transit rate relative to auto decreases.