The integration of transit-oriented development (TOD) and green urbanism has been increasingly recognized as a sustainable policy for the future development of cities. Although green amenities such as urban parks play a crucial role in promoting walkability and enhancing urban vitality, the causal impacts within TOD are unclear. In particular, it remains to be seen whether the observed increases in urban vitality are directly attributable to parks or are instead driven by other factors such as population density and land use. To elucidate this point, this study investigates the causal effects of neighborhood parks on visiting populations around railway stations in the Tokyo metropolitan area. Using mobile phone location data to capture staying-based activities, we first apply the propensity score matching method to estimate the impact of neighborhood parks on staying-based activity around stations and then employ a causal forests model to clarify how different locational characteristics influence park effectiveness. Our results show that neighborhood parks significantly increase visiting populations, with stronger effects on weekends than on weekdays. Furthermore, the positive impacts are more pronounced in areas with high densities of commercial facilities and greater transport accessibility. These findings provide empirical evidence on the role of urban parks in enhancing vibrant environments, offering policy-relevant insights for integrated land use and transportation planning, particularly in TOD.
The combined use of metro and dockless bike-sharing (M-DBS) is an emerging strategy for improving public-transport efficiency and tackling first-/last-mile gaps, and it is essential for sustainable urban mobility. However, existing studies often overlook land-use pattern heterogeneity and the nonlinear impacts of the built environment on M-DBS use. This study, using Shanghai as a case, employs K-Means clustering to identify four land-use patterns around metro stations—low-density, mixed-use, transit hub, and commercial core—and applies the XGBoost algorithm to model the nonlinear relationship between the built environment and M-DBS use across these patterns. We find that the built environment factors influencing M-DBS use vary by land-use pattern. In particularly, bus accessibility, population density, the proportion of shopping facilities, and subway accessibility play key roles in the four land-use patterns, respectively. Furthermore, the relationship between built environment variables and M-DBS use is nonlinear, with varying thresholds and even reversed correlations in different patterns. For example, in the transit hub pattern, M-DBS use negatively correlates with the proportion of dining facilities, while in the commercial core pattern, the correlation is positive. This study also reaffirms the differing impacts of the built environment on the two feeder modes of M-DBS use. Based on these findings, we provide strategic recommendations to optimize the integration of dockless bike-sharing and metro systems, improving public transportation efficiency and supporting sustainable urban development.
"Friction of distance" in human geography holds that increasing separation between locations within the built environment imposes costs in the form of money, time, effort, and negative externalities such as stress and crime. This study considered the ability of land use and roadway travel characteristics together as a viable way to estimate the known relationship of distance in terms of suburbanization with violent crime measured by murder rates in primary cities for 147 of the largest U.S. metropolitan statistical areas (MSAs). Methodologies consisted of multivariable cross-sectional ordinary least squares (OLS) linear and quantile regression. Development of multiple models revealed that MSA-weighted population density consistently held statistically significant inverse relationships with primary city murder rates; elasticities ranged from -0.18 to -0.31 at the mean. The proportion of residents near city centers also had similar associations at smaller elasticities. Respective measures of urban area freeway lane kilometers per 100,000 primary city population and urban area per capita vehicle kilometers traveled each as sole suburbanization independent variables had elasticities of 0.12 and 0.40 without substantively compromising model strength. Average annual hours of delay per auto commuter and number of vehicles per primary city square kilometers had similar elasticities while increasing prognostic accuracy of certain models. Statistical significance of the suburbanization independent variables was as high as eight out of nine quantiles. Extenuating support is given to theories of urbanism, metropolitan expansion, spatial mismatch, and social disorganization as friction of distance exacerbates detrimental inner-city conditions known to facilitate violent crime.
The location decision of knowledge-intensive firms (KIF) depends on many factors. In a globally connected knowledge economy, long-distance accessibility is one such factor, which is provided by long-distance railway, inter alia. Areas in proximity to long-distance rail stations are therefore candidates for the settlement of KIF. At the same time, the concept of Transit-Oriented Development (TOD) posits the development of areas around public transit hubs for residential and commercial use with high densities and diversity of uses, which are additional factors for the settlement of KIF as they increase the location attractiveness of workplaces for employees. In Germany, TOD has not yet been embraced for guidance in actual planning procedures, at least not in an outspoken sense. However, we hypothesize that effects are likely to be detected: rail station areas that exhibit more pronounced TOD characteristics will be associated with higher counts of KIF. To test the hypothesis, we develop and apply a quantitative operationalization of TOD and contrast it with the settlement of KIF at 178 long-distance rail stations in Germany. The results show that the relationships between the variables are as hypothesized. Higher degrees of TOD implementation are associated with higher counts of KIF. Therefore, as our methodological innovation, we suggest adding depth of knowledge, here capturing the absolute number of KIF, as a TOD outcome dimension variable besides the conventional TOD dimensions. These results have policy relevance for (German) planning procedures as they show that more active use of TOD implementation practices is likely to entail positive feedback loops for cities by attracting the knowledge economy.
Accessibility, a transport and land-use performance metric, is an umbrella term for several methodological approaches quantifying access to opportunities that can impact travel behavior. Some accessibility measures are easy to estimate and interpret, although they are based on restrictive assumptions, while others are more realistic but impose higher requirements for inputs, estimation, and interpretation. Selecting the measure of accessibility to incorporate in planning practice is a continuous challenge that professionals face. In this paper, we compare the strength of association between non-competitive and competition-based opportunity measures and the share of transit mode users at the Census Tract level of analysis in Toronto, Montreal, and Vancouver, Canada. Our findings confirm the positive association between all accessibility measures to jobs by public transit and transit mode share and identify that the simple cumulative opportunities calculated at the mean transit travel time of the region and gravity-based measures result in higher explanatory power compared to more complex competition-based measures, with a negligible level of difference between the former two. The insights from this paper can be of value to analysts and practitioners seeking to select accessibility measures that are representative of real conditions, easy to calculate and interpret, and with high predictive power of travel behavior in planning practice.
Data on private, off-street parking infrastructure are scarce, as the data are generally not systematically collected by public institutions and difficult to survey due to often inaccessible locations. As data are needed for targeted parking policies and for research related to parking, good estimates are required. Although such estimations are used in several studies, no study has yet compared different estimation approaches or evaluated their accuracy. Based on a literature review, we develop four approaches to estimating residential private parking supply and apply them to the city of Aachen, Germany. By comparing the estimates to a manually conducted survey in diverse neighborhoods, we evaluate and discuss the accuracy of the approaches. Our results highlight the difficulty of obtaining reliable data on private parking spaces, as the estimates vary considerably. Among the tested approaches, the application of a binary logit model based on real estate data provides the closest match to the surveyed parking supply. As this approach is capable of accounting for the large spatial variety in parking space availability within a city, it is not only suitable for city-wide estimations, but also for estimating the residential private parking supply in smaller spatial units like city blocks.
This study presents vehicle ownership simulation (VOSim) within the STELARS (Simulator for Transportation, Energy, LAnd use for Regional System) framework. VOSim follows an event-based decision process adopting a hybrid of continuous and discrete time simulation techniques. In STELARS, each household agent subscribes to a list of events (e.g., childbirth) that makes the agent actively adjust his or her vehicle fleet. Being active, agents make two interconnected decisions: vehicle transaction and type choice. In the vehicle transaction stage, for households that never owned a vehicle, the timing of the first vehicle purchase decision is simulated. For households with vehicles, their decision to add, dispose, or replace a vehicle is simulated. In the vehicle type choice stage, an agent’s decision to choose vehicles by body, vintage, fuel, and technology type is simulated. Vehicle transaction is simulated as a continuous-time decision using a hazard-based model. Once the timing of the transaction is determined, the vehicle type choice simulation transitions into a discrete-time step. This paper reports VOSim predictions and multi-year validation for the Okanagan region in Canada for the 2011-2021 period. Multi-year validation results confirm a satisfactory accuracy level. Prediction results suggest that a higher proportion of first-time vehicle purchasers reside in areas with lower accessibility to transit. A higher share of suburban dwellers is predicted to own alternative fuel vehicles. Overall, the VOSim adds capacity to integrated urban models to simulate vehicle ownership using a behaviorally realistic simulation procedure and be sensitive to plans and policies through an equity lens, such as who makes the first vehicle purchase decision.
Accessibility describes the potential to reach opportunities and is widely used to assess the ease of reaching destinations through urban transport systems. Although much attention has been given to investigate bicycle-accessibility and metro-accessibility methods, extending these methods to model bicycle-metro integration travel at the city scale remains challenging. Based on Hansen’s accessibility model, this study proposes three different models to measure bicycle accessibility within metro catchment areas. In particularly, key factors such as trip purposes, bicycle suitability, total travel time, and traffic demand are incorporated into the accessibility models. These proposed models have been tested and compared using empirical data from Shanghai. Overall, metro stations with multiple interchange lines, cycling-friendly facilities and diverse surrounding activities tend to exhibit higher bicycle accessibility, particularly those located in the city center. For areas with low bicycle accessibility in the city, such as Baoshan Road Station and Anshan Xincun Station, targeted improvement measures can be implemented to enhance bicycle-metro integration and bicycle accessibility.
Parking plays a vital role in shaping land use and transport. Despite occupying significant portions of urban space, detailed data on parking locations and capacities are often unavailable. Recognizing the critical significance of such data for comprehensive transportation modeling and sustainable urban planning, this study presents two statistical models designed to predict the available on-street parking length in urban traffic analysis zones. The first model uses OpenStreetMap (OSM) data as its primary input, while the second is based on official parking inventory data from the city of Berlin. Both models are built using multiple linear regression, with land use and built environment characteristics as independent variables. The models are evaluated by applying them to the city of Munich. This research provides new insights into the spatial distribution of urban on-street parking and offers a practical approach for estimating parking supply to support sustainable urban development strategies.
This study investigates parcel delivery patterns in Belo Horizonte, Brazil, to elucidate the influence of spatial inequalities, urban structure, socioeconomic factors, and retail diversity on delivery demand, employing spatial regression models. The results reveal that income and retail diversity positively impact parcel delivery, while food deserts drive increased reliance on e-commerce due to limited local options. In particular, the distance from the city center negatively affects delivery patterns, highlighting spatial inequities. Areas characterized by social inequalities exhibit greater delivery activity, highlighting e-commerce as a vital alternative where local services are scarce. These findings advocate for integrated urban planning policies that strategically leverage parcel delivery services to achieve more equitable access, address service gaps, and foster delivery expansion in marginalized areas.
Accurately measuring street dimensions is essential to evaluating how their design influences both travel behavior and safety. However, gathering street-level information at city-scale with precision is difficult given the quantity and complexity of urban intersections. To address this challenge in the context of pedestrian crossings — a crucial component of walkability — we introduce a scalable and accurate method for automatically measuring crossing distance at both marked and unmarked crosswalks, applied to America’s 100 largest cities. First, OpenStreetMap coordinates were used to retrieve satellite imagery of intersections throughout each city — totaling roughly three million images. Next, Meta’s Segment Anything Model was trained on a manually labelled subset of these images to differentiate drivable from non-drivable surfaces (i.e., roads vs. sidewalks). Third, all available crossing edges from OpenStreetMap were extracted. Finally, crossing edges were overlaid on the segmented intersection images, and a grow-cut algorithm was applied to connect each edge to its adjacent non-drivable surface (e.g., sidewalk, private property, etc.), thus enabling the calculation of crossing distance. This achieved 93% accuracy in measuring crossing distance, with a median absolute error of 2 feet 3 inches (0.69 meters), when compared to manually verified data for an entire city. Across the 100 largest U.S. cities, median crossing distances ranged from 32 feet to 78 feet (9.8 – 23.8m), with detectable regional patterns. Median crossing distance also displayed a positive relationship with the cities’ year of incorporation, illustrating in a novel way how American city planning increasingly emphasizes wider (and more car-centric) streets. These findings identified opportunities to improve pedestrian safety and increase walkability at multiple scales, from the individual block to the entire city.
The rational inattention theory aims to evaluate instances in which a decision is made in an information-rich environment where consumers cannot process all information due to limited cognitive capacity. In contrast to classical random utility-maximizing models, rational inattention discrete choice models do not assume that decision-makers make choices with complete knowledge of the alternatives. Today's information technology tools create a decision-making environment in which information is plentiful and easily accessible. Yet, it is cognitively impossible for households to be aware of every aspect of available options. This study uses rational inattention theory to investigate residential location choices in the Greater Toronto Area (GTA) during the COVID-19 pandemic, using an efficient-adaptive stated preference dataset collected in July 2021. The rational inattention theory requires identifying information processing costs and marginal probabilities as decision-makers' prior beliefs. The empirical model of this paper proposes using the time respondents spend on choice problems to measure their attention span and the latent preferences produced from the efficient-adaptive survey to measure their prior beliefs.
The limitation of land causes conflicts over how it is used, particularly in functional urban areas. Although Not in My Backyard (NIMBY) effects are often associated with established residential neighborhoods, they may also arise in rapidly developing suburban and peri-urban zones, where transport investments intersect with existing settlements and ongoing land-use change. Due to their specific nature, projects for new railway lines are more likely to receive a negative reception, because they can adversely affect property values, cause buyouts and expropriations, and introduce significant changes to a settlement’s spatial layout. The feasibility of different infrastructure project variants could be affected by varying levels of risk. Therefore, the occurrence of the NIMBY phenomenon necessitates a precise, standardized approach to its assessment. For this reason, the main aim of this study is to develop an algorithm to assess the scale of NIMBY responses for railway construction investments in densely populated Functional Urban Area (FUA) commuting zones, involving multicriteria spatial analysis. The algorithm is tested in the selected Łomża FUA railway construction project located in northeastern Poland. The project involves both the restoration of the railway line and its construction in a new location. This study highlights the analyses of spatial structural change that are crucial for understanding local resistance, since in suburbanizing areas, the intensity of NIMBY is strongly correlated with demographic shifts and patterns of land-use transformation. The Łomża FUA case confirms this.
This study explores how different aspects of walkability are associated with residential and retail gentrification in U.S. cities. Using data from the American Community Survey (ACS) and the U.S. Environmental Protection Agency’s Smart Locations Dataset (SLD), we examine walkability scores along with their underlying factors including the diversity of amenities, proximity to transit, and intersection density to predict gentrification. The ACS provides detailed demographic, socioeconomic, and housing data at the neighborhood level, enabling analysis of population shifts and economic changes over time. The SLD offers spatial indicators of walkability based on consistent national methodologies, making it a valuable tool for comparing built-environment characteristics across cities. Our findings show that overall walkability and neighborhood amenities are positively associated with both residential and retail gentrification. In addition, higher intersection density is linked to residential gentrification, underscoring the importance of neighborhood connectivity in attracting higher-income residents. These results highlight the complexity of gentrification and the need for more targeted policy interventions that address the various components of walkability and connectivity.
Children's active transport to school has multiple health, social, economic, and environmental benefits, and the literature on ways to support children’s active accessibility is vast. The consistent conclusion from this research is that the distance between home and school is a key determinant of whether a child will walk or cycle to and from the school gate. While distance is undoubtedly of central importance to the active school travel puzzle, our understanding of children’s sensitivity to increases in distance remains nascent. How far is too far when it comes to active school access? Using survey data from 6,629 school students in Australia, this paper explores this question through a nuanced focus on the sensitivity of active transport to school (ATS) to changes in trip distance. More specifically, a multinomial logistic regression model is used to analyze the heterogeneity of elasticity to distance, and the nature of the relationship between distance elasticity and land-use and demographic segments. The findings confirm existing understandings that distance and local land use are significant factors associated with ATS. By using a piecewise treatment of distance and estimating point elasticities, the model also shows that mode-choice sensitivity to distance varies across places and populations and is itself non-linear. The turning point of the multinomial logistic regression function with respect to distance is between 1 km and 3 km, indicating that a percentage increase in distance within this range is most likely to deter active school travel. This novel finding provides much-needed clarity to existing understandings of the sensitivity of ATS to distance. Such understandings are central to policy aspirations seeking to design school catchments with active accessibility as a desired outcome.
Accessibility to public open space (POS) is critical for promoting walking and outdoor physical activity. In dense, vertically developed cities, traditional two-dimensional accessibility measures fail to capture the complexities of vertical and slope movement. This study addressed this gap by incorporating three-dimensional (3D) distance into walking accessibility assessment and examining spatial equality across multi-type POSs. A research framework was established integrating Gini index-based equality measurement, regression analysis of influencing factors, and multivariate grouping of street blocks to inform planning strategies for equitable POS distribution. The framework was applied to the case of the Kowloon district in the vertical city of Hong Kong. Spatial (in)equality of 3D walking accessibility for different POSs, including local POSs, district and regional POSs, POSs with children’s playgrounds and POSs with fitness equipment for the elderly was measured and interpreted. The results suggested that small-scale local POSs were undersupplied compared to large ones, while accessibility inequality was more pronounced for child- and elderly-oriented facilities. Ethnicity was the most influential social-economic factor in the spatial heterogeneity of POS distribution. Spatial grouping analysis identified only 16.03% of the area as "optimizable zones" suitable for renewal interventions, with the remainder classified as "deadlock" (high-density constraints) or "stagnant" (low renewal urgency) zones. Future urban renewal projects should pay more attention to the improvement of POS quality and spatial equality. The proposed framework offers transferable methodologies for enhancing POS spatial justice in high-density cities through data-informed planning and governance.
Transportation research has extensively examined the influence of both the built and natural environment on active travel. While most studies assume linear relationships, some evidence indicates that this might not always be the case. This paper addresses this by identifying the nature of the relationship between the built and natural environment (BNE) and active travel (AT) across several trip purposes: school, shopping, work, and leisure trips in Montreal, Canada. We also identify areas with low and high potential for active travel. Using Generalized Linear Models with the Tweedie family and including a spatial lag covariate, we found that the relationship between BNE and AT is not always linear. In some cases, higher access levels to sidewalks, bike lanes, walkable destinations, and transit stops, constantly increase AT but with cubic or logarithmic relationships. Other variables, such as dwelling density, intersection density, park access, tree coverage, industrial diversity, and proximity to water bodies, also encourage active travel but only up to a certain threshold, beyond which further increases do not increase AT, and in some cases, can lead to a decline, forming an inverted "U" relationship. These relationships vary across trip purposes. Central areas in Montreal show the best potential to support active travel, while the rest of the city displays low levels of support, depending on the trip's purpose. The findings highlight the importance of accounting for non-linear relationships, as improvements in the BNE do not always translate into higher levels of active travel.
Interjurisdictional commuting is increasingly prevalent in China, yet the socioeconomic and mobility disparities between interjurisdictional and intrajurisdictional commuters remain underexamined. This study investigates the socioeconomic, residential, and travel-related differences between interjurisdictional and intrajurisdictional commuters in four cities—Guangzhou, Shenzhen, Foshan, and Dongguan—using cellphone data of 15.2 million users on October 18, 2023. As commuters are not randomly assigned across jurisdictional boundaries, propensity score matching was employed to adjust for differences in workplace characteristics before comparing the two groups. The analysis reveals that interjurisdictional commuters are younger, more likely to be male and migrants, and tend to live in low-rent neighborhoods with poorer access to urban amenities. They also experience significantly longer commuting distances, durations, and higher transport costs compared to their intrajurisdictional counterparts. These disparities reflect not only individual choices but also structural challenges, such as housing affordability gaps, fragmented transport governance, and insufficient regional planning. The study contributes to the spatial mismatch and transport poverty literature by highlighting the regional dimension of commuting inequalities in polycentric urban systems. It underscores the need for integrated transit planning in peripheral cities, and targeted support for long-distance commuters. These findings offer policy-relevant insights for fostering equitable mobility in rapidly urbanizing regions.
This paper examines the role of telework adoption and preferences in residential relocation, focusing on how residential dissonance influences the intention to move. Specifically, if individuals more positive toward telework find their current residential location mismatched with their preferences, they are more likely to experience residential dissonance. This dissatisfaction could drive them to relocate to more suburban areas, contributing to urban sprawl. In this context, the 15-minute city could play a crucial role in the decision, as individuals may be drawn to urban environments with these characteristics, potentially mitigating the pressure for suburban relocation. A Structural Equation Model is estimated to test the study hypothesis using data from an online survey conducted in the Lisbon Metropolitan Area during spring and summer 2021. The results suggest that while attitudes toward telework do not significantly influence the decision to move, teleworking practices, particularly experiences during the pandemic do, indicating that telework could contribute to suburbanization. Nevertheless, the variables that capture the push factors for residential mobility (being young, living in a small house, having children, and being a renter) have a more decisive influence on the decision to move than the effects of telework, residential satisfaction, and residential dissonance. As for living in a 15-minute city, individuals residing in areas more aligned with the concept are less satisfied with the quality of public space, although more satisfied with accessibility levels. Living in a 15-minute city has a positive effect on the intention to move. This suggests that the impact of living in a 15-minute city may not be as straightforward as often assumed, highlighting the complexities of residential satisfaction in these environments.
This paper presents a data-driven agent-based model that simulates the weekly grocery shopping behavior of disadvantaged consumers in highly segregated lower eastside neighborhoods of Detroit, Michigan. We focus on neighborhoods experiencing severe disinvestment to analyze the shopping behavior of residents after all major regional and national supermarket chains abandoned the city. The presented model is unique in that it utilizes detailed shopping behavior data collected to examine travel in marginalized communities, specifically among residents in severe poverty who are often overlooked in the travel behavior literature. The research shows that in extreme socio-economic decline, sociodemographic variables (such as class) can become more relevant than the built environment (land-use mix, density, and street connectivity) in determining access and influencing mobility. After identifying unique groups of household agents, we design rules that utilize probability distributions generated from survey responses. The decision-making of agents that emulate households is habitual rather than utility driven. Modeled behavior is designed based on stated preferences, which may contradict premises such as the “shortest distance to the nearest shop” approach, a common assumption in the literature. We also report on three what-if scenarios to evaluate how major population changes would affect the results.
