Analysis of Metro ridership at station level and station-to-station level in Nanjing: an approach based on direct demand models

A growing base of research adopts direct demand models to reveal associations between transit ridership and influence factors in recent years. This study is designed to investigate the factors affecting rail transit ridership at both station level and station-to-station level by adopting multiple regression model and multiplicative model respectively, specifically using an implemented Metro system in Nanjing, China, where Metro implementation is on the rise. Independent variables include factors measuring land-use mix, intermodal connection, station context, and travel impedance. Multiple regression model proves 11 variables are significantly associated with Metro ridership at station level: population, employment, business/office floor area, CBD dummy variable, number of major educational sites, entertainment venues and shopping centers, road length, feeder bus lines, bicycle park-and-ride (P&R) spaces, and transfer dummy variable. Results from multiplicative model indicate that factors influencing Metro station ridership may also influence Metro station-to-station ridership, varied by both trip ends (origin/destination) and time of day. In comparison with previous case studies, CBD dummy variable and bicycle P&R are statistically significant to explain Metro ridership in Nanjing. In addition, Metro travel impedance variables have significant influence on station-to-station ridership, representing the basic time-decay relationship in travel distribution. Potential implications of the model results include estimating Metro ridership at station level and station-to-station level by considering the significant variables, recognizing the necessity to establish a cooperative multi-modal transit system, and identifying opportunities for transit-oriented development.     

Examining the relationship between built environment and metro ridership at station-to-station level

Very few studies have examined the impact of built environment on urban rail transit ridership at the station-to-station (origin-destination) level. Moreover, most direct ridership models (DRMs) tend to involve simple a prior assumed linear or log-linear relationship in which the estimated parameters are assumed to hold across the entire data space of the explanatory variables. These models cannot detect any changes in the linear (or non-linear) effects across different values of the features of built environment on urban rail transit ridership, which possibly induces biased results and hides some non-negligible and detailed information. Based on these research gaps, this study develops a time-of-day origin-destination DRM that uses smart card data pertaining to the Nanjing metro system, China. It applies a gradient boosting regression trees model to provide a more refined data mining approach to investigate the non-linear associations between features of the built environment and station-to-station ridership. Data related to the built environment, station type, demographics, and travel impedance including a less used variable – detour, were collected and used in the analysis. The empirical results show that most independent variables are associated with station-to-station ridership in a discontinuous non-linear way, regardless of the time period. The built environment on the origin side has a larger effect on station-to-station ridership than the built environment on the destination side for the morning peak hours, while the opposite holds for the afternoon peak hours and night. The results also indicate that transfer times is more important variables than detour and route distance.     

Impacts of weather on public transport ridership: Results from mining data from different sources

The existing studies concerning the influence of weather on public transport have mainly focused on the impacts of average weather conditions on the aggregate ridership of public transit. Not much research has examined these impacts at disaggregate levels. This study aims to fill this gap by accounting for intra-day variations in weather as well as public transport ridership and investigating the effect of weather on the travel behavior of individual public transit users. We have collected smart card data for public transit and meteorological records from Shenzhen, China for the entire month of September 2014. The data allow us to establish association between the system-wide public transit ridership and weather condition on not only daily, but also hourly basis and for each metro station. In addition, with the detailed trip records of individual card holders, the travel pattern by public transit are constructed for card holders and this pattern is linked to the weather conditions he/she has experienced. Multivariate modeling approach is applied to analyze the influence of weather on public transit ridership and the travel behavior of regular transit users. Results show that some weather elements have more influence than others on public transportation. Metro stations located in urban areas are more vulnerable to outdoor weather in regard to ridership. Regular transit users are found to be rather resilient to changes in weather conditions. Findings contribute to a more in-depth understanding of the relationship between everyday weather and public transit travels and also provide valuable information for short-term scheduling in transit management.     

Examining determinants of rail ridership: a case study of the Orlando SunRail system

The current study contributes to the literature on transit ridership by considering daily boarding and alighting data from a recently launched commuter rail system in Orlando, Florida – SunRail. The analysis is conducted based on daily boarding and alighting data for 10 months for the year 2015. With the availability of repeated observations for every station, the potential impact of common unobserved factors affecting ridership variables are considered. The current study develops an estimation framework, for boarding and alighting separately, that accounts for these unobserved effects at multiple levels – station, station-week and station-day. In addition, the study examines the impact of various observed exogenous factors such as station level, transportation infrastructure, transit infrastructure, land use, built environment, sociodemographic and weather variables on ridership. The model system developed will allow us to predict ridership for existing stations in the future as well as potential ridership for future expansion sites.     

Evaluating light rail sketch planning: actual versus predicted station boardings in Phoenix

In recent years, transit planners are increasingly turning to simpler, faster, and more spatially detailed “sketch planning” or “direct demand” models for forecasting rail transit boardings. Planners use these models for preliminary review of corridors and analysis of station-area effects, instead of or prior to four-step regional travel demand models. This paper uses a sketch-planning model based on a multiple regression originally fitted to light-rail ridership data for 268 stations in nine U.S. cities, and applies it predictively to the Phoenix, Arizona light-rail starter line that opened in December, 2008. The independent variables in the regression model include station-specific trip generation and intermodal–access variables as well as system-wide variables measuring network structure, climate, and metropolitan-area factors. Here we compare the predictions we made before and after construction began to pre-construction Valley Metro Rail predictions and to the actual boardings data for the system’s first 6 months of operations. Depending on the assumed number of bus lines at each station, the predicted total weekday ridership ranged from 24,767 to 37,907 compared with the average of 33,698 for the first 6 months, while the correlation of predicted and observed station boardings ranged from r = 0.33 to 0.47. Sports venues, universities, end-of-line stations, and the number of bus lines serving each station appear to account for the major over- and under-predictions at the station level.     

A comparative study of aggregate and disaggregate gravity models using Seoul metropolitan subway trip data

Abstract

Existing origin constrained and doubly constrained gravity models have not been compared, theoretically or empirically, in terms of their forecasting power. Due to the newly advanced technology of intelligent transport systems, the expanded data presently available have made various models more comparable in terms of forecasting power. This paper uses archived automatic passenger counting (APC) data for urban rail in the Seoul metropolitan area. The APC data contains information about each trip's origin, destination, ticket type, fare, and distance on a daily basis. The objective of this paper is to compare the goodness-of-fit of aggregate and disaggregate gravity modeling using these data. A Hyman aggregate gravity model is used as the aggregate model without the spatial effect. The disaggregate model adopts a multinomial logit as the destination choice model with the spatial effect. In general, while the formulation of aggregate and disaggregate gravity model models are similar, the calibration and parameter estimation methods of the two models are different. As a result, this empirical study demonstrates that the variation in goodness-of-fit and forecasting power largely depends on the estimation method and selected variables. The forecasting power of the disaggregate modeling approach outperforms that of the aggregate model. This paper further confirms that spatial arrangement plays important roles in gravity modeling.     

Spatio-temporal analysis of rail station ridership determinants in the built environment

The development of new routes and stations, as well as changes in land use, can have significant impacts on public transit ridership. Thus, transport departments and governments should seek to determine the level and spatio-temporal dependency of these impacts with the aim of adjusting services or improving planning. However, existing studies primarily focus on predicting ridership, and pay relatively little attention to analyzing the determinants of ridership from temporal and spatial perspectives. Consequently, no comprehensive cognition of the spatio-temporal relationship between station ridership and the built environment can be obtained from previous models, which makes them unable to facilitate the optimization of transportation demands and services. To rectify this problem, we have employed a Bayesian negative binomial regression model to identify the significant impact factors associated with entry/exit ridership at different periods of the day. Based on this model, we formulated geographically weighted models to analyze the spatial dependency of these impacts over different periods. The spatio-temporal relationship between station ridership and the built environment was analyzed using data from Beijing. The results reveal that the temporal impacts of most ridership determinants are related to the passenger trip patterns. Furthermore, the spatial impacts correspond with the determinants’ spatial distribution, and the results give some implications on urban and transportation planning. This analysis gives a common analytical framework analyzing impacts of urban characteristics on ridership, and extending researches on how we capture the impacts of urban and other factors on ridership from a comprehensive perspective.

     

Quantifying the relative contribution of factors to household vehicle miles of travel

Household vehicle miles of travel (VMT) has been exhibiting a steady growth in post-recession years in the United States and has reached record levels in 2017. With transportation accounting for 27 percent of greenhouse gas emissions, planning professionals are increasingly seeking ways to curb vehicular travel to advance sustainable, vibrant, and healthy communities. Although there is considerable understanding of the various factors that influence household vehicular travel, there is little knowledge of their relative contribution to explaining variance in household VMT. This paper presents a holistic analysis to identify the relative contribution of socio-economic and demographic characteristics, built environment attributes, residential self-selection effects, and social and spatial dependency effects in explaining household VMT production. The modeling framework employs a simultaneous equations model of residential location (density) choice and household VMT generation. The analysis is performed using household travel survey data from the New York metropolitan region. Model results showed insignificant spatial dependency effects, with socio-demographic variables explaining 33 percent, density (as a key measure of built environment attributes) explaining 12 percent, and self-selection effects explaining 11 percent of the total variance in the logarithm of household VMT. The remaining 44 percent remains unexplained and attributable to omitted variables and unobserved idiosyncratic factors, calling for further research in this domain to better understand the relative contribution of various drivers of household VMT.     

A multiobjective planning model for intercity train seat allocation

This paper presents a multiobjective planning model for generating optimal train seat allocation plans on an intercity rail line serving passengers with many‐to‐many origin‐destination pairs. Two planning objectives of the model are to maximise the operator's total passenger revenue and to minimise the passenger's total discomfort level. For a given set of travel demand, train capacity, and train stop‐schedules, the model is solved by fuzzy mathematical programming to generate a best‐compromise train seat allocation plan. The plan determines how many reserved and non‐reserved seats are to be allocated at each origin station for all subsequent destination stations on each train run operated within a specified operating period. An empirical study on the to‐be‐built Taiwan's high‐speed rail system is conducted to demonstrate the effectiveness of the model. The model can be used for any setting of travel demand and stop‐schedules with various train seating capacities.     

Activity imputation for trip-chains elicited from smart-card data using a continuous hidden Markov model

Although smart-card data were expected to substitute for conventional travel surveys, the reality is that only a few automatic fare collection (AFC) systems can recognize an individual passenger's origin, transfer, and destination stops (or stations). The Seoul metropolitan area is equipped with a system wherein a passenger's entire trajectory can be tracked. Despite this great advantage, the use of smart-card data has a critical limitation wherein the purpose behind a trip is unknown. The present study proposed a rigorous methodology to impute the sequence of activities for each trip chain using a continuous hidden Markov model (CHMM), which belongs to the category of unsupervised machine-learning technologies. Coupled with the spatial and temporal information on trip chains from smart-card data, land-use characteristics were used to train a CHMM. Unlike supervised models that have been mobilized to impute the trip purpose to GPS data, A CHMM does not require an extra survey, such as the prompted-recall survey, in order to obtain labeled data for training. The estimated result of the proposed model yielded plausible activity patterns that are intuitively accountable and consistent with observed activity patterns.     

Evaluation of a public transportation level of service concept

The level of service of a bus line is evaluated by its operational characteristics, particularly by the ratio between average bus travel time on a given route and the average passenger car travel time on the shortest distance between the origin and the destination of the bus in question. It is shown that the level-of-service measure may be predicted by such independent variables as route length, average distance between bus stations, number of signalized and unsignalized intersections, and the ratio between such intersections. It is hypothesized that use of other independent variables such as boarding and alighting passengers, or volume to capacity ratio on the route concerned, could improve the predictive power of the suggested models. Further research is recommended on the effect of these latter variables and other operational variables which might influence bus level of service, and also on the comparison between direct bus lines and lines which use transfer points.     

A combined destination and route choice model for a bicycle sharing system

This paper studies the supply variables that influence the destination and route choices of users of a bicycle sharing system in the Chilean city of Santiago. A combined trip demand logit model is developed whose explanatory variables represent attributes relating to the topology of the possible routes and other characteristics such as the presence of bikeways, bus service and controlled intersections. The data for the explanatory variables and system users were collected through field surveys of the routes and interviews conducted at the system stations. The results of the model show that proximity to stops on the Santiago Metro and the existence of bikeways are the main factors influencing destination and route choices. Also indicated by the model estimates are gender differences, a preference for tree-lined routes and an avoidance of routes with bus services. Finally, the outcomes reveal considerable potential for the integration of bicycle sharing systems with Metro transit.     

Impact evaluation of a mass transit fare change on demand and revenue utilizing smart card data

Transit fares are an effective tool for demand management. Transit agencies can raise revenue or relieve overcrowding via fare increases, but they are always confronted with the possibility of heavy ridership losses. Therefore, the outcome of fare changes should be evaluated before implementation. In this work, a methodology was formulated based on elasticity and exhaustive transit card data, and a network approach was proposed to assess the influence of distance-based fare increases on ridership and revenue. The approach was applied to a fare change plan for Beijing Metro. The price elasticities of demand for Beijing Metro at various fare levels and trip distances were tabulated from a stated preference survey. Trip data recorded by an automatic fare collection system was used alongside the topology of the Beijing Metro system to calculate the shortest path lengths between all station pairs, the origin–destination matrix, and trip lengths. Finally, three fare increase alternatives (high, medium, and low) were evaluated in terms of their impact on ridership and revenue. The results demonstrated that smart card data have great potential with regard to fare change evaluation. According to smart card data for a large transit network, the statistical frequency of trip lengths is more highly concentrated than that of the shortest path length. Moreover, the majority of the total trips have a length of around 15 km, and these are the most sensitive to fare increases. Specific attention should be paid to this characteristic when developing fare change plans to manage demand or raise revenue.     

Comparing decision tree algorithms to estimate intercity trip distribution

Traditional trip distribution models usually ignore the fact that destination choices are made individually in addition to aggregated factors, such as employment and average travel costs. This paper proposes a disaggregated analysis of destination choices for intercity trips, taking into account aggregated characteristics of the origin city, an impedance measurement and disaggregated variables related to the individual, by applying nonparametric Decision Tree (DT) algorithms. Furthermore, each algorithm’s performance is compared with traditional gravity models estimated from a stepwise procedure (1) and a doubly constrained procedure (2). The analysis was based on a dataset from the 2012 Origin-Destination Survey carried out in Bahia, Brazil. The final selected variables to describe the destination choices were population of the origin city, GDP of the origin city and travel distances at an aggregated level, as well as the variables: age, occupation, level of education, income (monthly), number of cars per household and gender at a disaggregated one. The comparison of the DT models with gravity models demonstrated that the former models provided better accuracy when predicting the destination choices (trip length distribution, goodness-of-fit measures and qualitative perspective). The main conclusion is that Decision Tree algorithms can be applied to distribution modeling to improve traditional trip distribution approaches by assimilating the effect of disaggregated variables.     

Exploring the link between the neighborhood typologies,bicycle infrastructure and commuting cycling over time and the potential impact on commuter GHG emissions

This paper investigates the evolution of urban cycling in Montreal, Canada and its link to both built environment indicators and bicycle infrastructure accessibility. The effect of new cycling infrastructure on transport-related greenhouse gas (GHG) emissions is then explored. More specifically, we aim at investigating how commuting cycling modal share has evolved across neighborhood built-environment typologies and over time in Montreal, Canada. For this purpose, automobile and bicycle trip information from origin–destination surveys for the years 1998, 2003 and 2008 are used. Neighborhood typologies are generated from different built environment indicators (population and employment density, land use diversity, etc.). Furthermore, to represent the commuter mode choice (bicycle vs automobile), a standard binary logit and simultaneous equation modeling approach are adopted to represent the mode choice and the household location. Among other things, we observe an important increase in the likelihood to cycle across built environment types and over time in the study region. In particular, urban and urban-suburb neighborhoods have experienced an important growth over the 10 years, going from a modal split of 2.8–5.3% and 1.4–3.0%, respectively. After controlling for other factors, the model regression analysis also confirms the important increase across years as well as the significant differences of bicycle ridership across neighborhoods. A statistically significant association is also found between the index of bicycle infrastructure accessibility and bike mode choice – an increase of 10% in the accessibility index results in a 3.7% increase in the ridership. Based on the estimated models and in combination with a GHG inventory at the trip level, the potential impact of planned cycling infrastructure is explored using a basic scenario. A reduction of close to 2% in GHG emissions is observed for an increase of 7% in the length of the bicycle network. Results show the important benefits of bicycle infrastructure to reduce commuting automobile usage and GHG emissions.     

Walking to the bus: perceived versus actual walking distance to bus stops for older adults

The walking trip from an origin or destination to a bus stop or transit station can be a barrier to riding transit for older adults (over age 60) who may walk more slowly than others or experience declining physical mobility. This article examines the relationship between transit ridership and proximity to fixed-route transit stations using survey data for older adults in Buffalo and Erie County, New York. Demographic and socio-economic characteristics—including age, sex, race, income, possessing a driver’s license, frequency of leaving home, and personal mobility limitations—are tested but do not display, in bi-variate analysis, statistically significant differences for transit riders versus non-transit riders. However, features of the built environment—including distance (actual and perceived) between home and transit stop, transit service level, population density, number of street intersections, metropolitan location, and neighborhood crime (property and violent) rate—display statistically significant differences for transit riders versus non-transit riders. Both objective and perceived walking distances to access fixed-route transit show statistically significant differences between transit riders and non-transit riders. Average walking distance from home to transit for non-transit riders—who mostly live in suburbs—is three times greater than average walking distance between home and the nearest transit stop for transit riders—who mostly live in the central city. When asked how near a bus stop is to their homes, transit riders slightly overestimate the actual distance, while non-transit riders underestimate the distance.     

Secrets of success: assessing the large increases in transit ridership achieved by Houston and San Diego transit providers

This paper summarizes and updates the findings from an earlier study by the same authors of transit systems in Houston (all bus) and San Diego (bus and light rail). Both systems achieved unusually large increases in transit ridership during a period in which most transit systems in other metropolitan areas were experiencing large losses. Based on ridership models estimated using cross section and time series data, the paper quantifies the relative contributions of policy variables and factors beyond the control of transit operators on ridership growth. It is found that large ridership increases in both areas are caused principally by large service increases and fare reductions, as well as metropolitan employment and population growth. In addition, the paper provides careful estimates of total and operating costs per passenger boarding and per passenger mile for Houston's bus operator and San Diego's bus and light rail operators. These estimates suggest that the bus systems are more cost-effective than the light rail system on the basis of total costs. Finally, the paper carries out a series of policy simulations to analyze the effects of transit funding levels and metropolitan development patterns on transit ridership and farebox recovery ratio.     

A note on the consistent aggregation of nested logit demand functions

The present paper derives a set of rules allowing for the consistent aggregation of nested logit travel demand functions across origin and destination zones. Presented aggregation rules are derived for the case when the mode choice is performed conditional on destination choice. The derivation is based on the principles of consistency between aggregate and disaggregate travel demand models introduced by Sweet as well as upon the sampling theory.     

Nature and/or nurture? Analyzing the determinants of transit ridership across US urbanized areas

Public subsidy of transit services has increased dramatically in recent years, with little effect on overall ridership. Quite obviously, a clear understanding of the factors influencing transit ridership is central to decisions on investments in and the pricing and deployment of transit services. Yet the literature about the causes of transit use is quite spotty; most previous aggregate analyses of transit ridership have examined just one or a few systems, have not included many of the external, control variables thought to influence transit use, and have not addressed the simultaneous relationship between transit service supply and consumption. This study addresses each of these shortcomings by (1) conducting a cross-sectional analysis of transit use in 265 US urbanized areas, (2) testing dozens of variables measuring regional geography, metropolitan economy, population characteristics, auto/highway system characteristics, and transit system characteristics, and (3) constructing two-stage simultaneous equation regression models to account for simultaneity between transit service supply and consumption. We find that most of the variation in transit ridership among urbanized areas – in both absolute and relative terms – can be explained by factors outside of the control of public transit systems: (1) regional geography (specifically, area of urbanization, population, population density, and regional location in the US), (2) metropolitan economy (specifically, personal/household income), (3) population characteristics (specifically, the percent college students, recent immigrants, and Democratic voters in the population), and (4) auto/highway system characteristics (specifically, the percent carless households and non-transit/non-SOV trips, including commuting via carpools, walking, biking, etc.). While these external factors clearly go a long way toward determining the overall level of transit use in an urbanized area, we find that transit policies do make a significant difference. The observed range in both fares and service frequency in our sample could account for at least a doubling (or halving) of transit use in a given urbanized area. Controlling for the fact that public transit use is strongly correlated with urbanized area size, about 26% of the observed variance in per capita transit patronage across US urbanized areas is explained in the models presented here by service frequency and fare levels. The observed influence of these two factors is consistent with both the literature and intuition: frequent service draws passengers, and high fares drive them away.     

Estimating freeway route travel time distributions with consideration to time‐of‐day,inclement weather,and traffic incidents

This paper develops an efficient probabilistic model for estimating route travel time variability, incorporating factors of time‐of‐day, inclement weather, and traffic incidents. Estimating the route travel time distribution from historical link travel time data is challenging owing to the interactions among upstream and downstream links. Upon creating conditional probability function for each link travel time, we applied Monte Carlo simulation to estimate the total travel time from origin to destination. A numerical example of three alternative routes in the City of Buffalo shows several implications. The study found that weather conditions, except for snow, incur minor impact on off‐peak and weekend travel time, whereas peak travel times suffer great variations under different weather conditions. On top of that, inclement weather exacerbates route travel time reliability, even when mean travel time increases moderately. The computation time of the proposed model is linearly correlated to the number of links in a route. Therefore, this model can be used to obtain all the origin to destination travel time distributions in an urban region. Further, this study also validates the well‐known near‐linear relation between the standard deviation of travel time per unit distance and the corresponding mean value under different weather conditions. Copyright © 2016 John Wiley & Sons, Ltd.