Question:
It is argued that walking behavior is influenced by how „walkable“ our environment is, but there is little research into the association between walking and walkability, and no clear consensus on how to measure it. This proof of concept study proposes a replicable method to measure walkability by making surfaces from public domain data to explore the relationship between walkability and walking behavior in cohort studies. The presentation will focus on the methodology we used.
Methods:
We generated three metrics which can be applied to any location using OpenStreetMap, an open data mapping resource: 1. the density of public transport access within a 640 m walk; 2. the density of local services and retail outlets (also within 640 m); 3. a measure of how the area around the point space can be traversed on foot accounting for topographical impedance, such as physical barriers. The metrics were calculated on a hexagonal grid for seven cities: Augsburg, Bochum, Dortmund, Essen, Greifswald, Halle and Mülheim. Using Kriging, we interpolated three walkability surfaces for each city which can be used with GIS software systems. They are now being used to measure the association between self-reported walking behavior of residents who were subjects in epidemiological studies all of which captured walking and exercise data comparably.
Results:
The study shows that publicly available map data can be used to derive walkability metrics that can be correlated with data on people and households. Viewed in the context of the characteristics of the cities, these metrics appear to have face validity. However the analysis of how our measures affect self reported walking behavior in the cohorts has not yet been completed.
Discussion:
We have shown that we can capture characteristics of walkability in built environments in a replicable and comparable way, and thus make an important contribution to understanding the interaction between people, environment and their behavior.
