Active travel is enthusiastically promoted in the Western world due to its clear and demonstrated individual and collective benefits. While active travel has been shown to be associated with features of the built environment such as density and land-use mix, it is also associated with walking and cycling accessibility-which we designate as active accessibility. However, the measurement of active accessibility is not straightforward and it can represent significantly different features of the built environment. This paper presents an extensive review of published research that measures active accessibility. We classified the literature into four categories based on the methodology used: distance-based, gravity-based or potential, topological or infrastructure-based, and walkability and walk score-type measures. A fifth category was created to classify outliers consisting of distinct methodological approaches or hybrids of the four main categories. We argue that almost all of these methods have conceptual and computational limitations, and that there are inconsistencies in the use of concepts and terms. Furthermore, no sensitivity analysis was carried out on the selected parameters. We conclude by presenting some guidelines that might improve the value and clarity of active accessibility research, theory, and practice.

No abstract

Abstract:There is a vast literature on the relationship between built environment and travel, emphasizing the importance of built environment as a determinant of travel. However, the majority of studies focuses on the characteristics of origins and neglects the influence that the destination might have on travel, despite the already demonstrated importance of destinations to explain travel. In this paper, we test the relationship between residential and workplace built environment and the commuting pattern of staff and students of the University of Lisbon, a multi-campus university. Data was obtained through a dedicated travel survey, containing 1474 georeferenced individuals. Chi-square analyses were developed to analyze differences between staff and students and between different campuses. A logistic regression model was developed to explain car commuting, controlling for socio-demographic data. Two different models were developed for students and staff.Our results show the built environment and associated multimodal accessibility of the campuses are important explanatory variables of commuting. Free parking at the campus is crucial for car commuting, especially for students. These results emphasize the importance of measuring destinations as explanatory variables and promoting good urban integration of the campus in the city, increasing its multimodal accessibility.

No abstract

Pedestrian accessibility has been growing in importance as an urban planning objective. Assessing it with gravity-based or potential accessibility measures requires the selection of an impedance function in order to reflect the friction of distance. The choice of impedance function is crucial to pedestrian accessibility assessment due to the level of spatial data detail required and also because perceived distances differ from physical distances. Here, we measure and compare 20 gravity-based measures, varying the impedance function and associated parameters. Correlation analysis revealed a significant and strong correlation between the measures. Factor analysis extracted two groups of measures, differing mainly in their maximum cutoff travel distance, i.e. the distance at which the impedance function reaches zero. Spatial analysis revealed that all measures produce similar spatial results in terms of identifying high and low accessibility locations but different values for medium accessibility locations. Places located at between 200 and 400 m from an opportunity are especially sensitive to the impedance function used. We promote a cumulative–Gaussian approach to measure pedestrian accessibility, as it explicitly includes the travel tolerance concept and we found it to be the most robust measure in terms of data variability.