Modelling multimodal mode choice behaviour considering spatial variability in level of service attributes

This study formulates a multimodal mode choice model that reflects the multimodal nature of transit trips, including the presence of access and egress legs. The model takes the form of a two-level mixed multinomial logit, where the primary mode choice is modelled at the top level and the access and egress mode choices are modelled at the bottom level. Further, the modelling framework recognizes the spatial variability in thelevel-of-service (LOS) attributes of primary and access/egress modes due to imprecise location data of trip destinations. To do so, the LOS attributes are specified as stochastic explanatory variables, with their distributions empirically characterized a priori. In addition, the study evaluates a suitable aggregation for defining traffic analysis zones (TAZs) for mode choice modelling in Indian cities. The empirical analysis and modelling in this research are applied to a home-to-work commute mode choice survey (revealed preference) data collected in Bengaluru, India. The results indicate that models that consider measurement error in LOS attributes (due to spatial aggregation) offered better data fit and willingness-to-pay measures than the models that ignore measurement errors. If the analyst is unable to consider the measurement error due to computational reasons, empirical models with Census block groups (with an average area of 0.27 km2) as TAZs yielded better results than those with Census wards (with an average area of 3.5 km2). Therefore, the study recommends adopting Census block groups as TAZs for modelling travel demand in Indian cities. Finally, policy analyses conducted using the proposed model suggest that expanding public transit along with access-end and egress-end connectivity to make transit accessible to a large population would be an effective way to increase transit mode share in Bengaluru.