Creating a Road Network Analysis Layer with Travel Time Estimates using Open-source Data

Abstract

A main focus of health geography research is assessing the impacts of environmental exposures on health. Exposures, in this instance, are anything in the environment that has the potential to negatively affect or positively contribute to health outcomes. Access to fast food outlets and obesity (Pearce et al. 2009), gambling opportunities and gambling behaviours (Pearce et al. 2008), alcohol availability and crime (Day et al. 2012), greenspace access and mental health (Nutsford et al. 2013), and traffic pollution exposure by mode of transport (Kingham et al. 2013) are examples of the type of research conducted by health geographers. Exposures are often modelled by determining proximity via a road network, such as measuring access to health services (Beere and Brabyn 2006; Brabyn and Beere 2006). Distance alone is not always the most appropriate measure as the time required to travel two equal distances may vary. Travel time arguably provides a more consistent basis for comparing exposures, however, the creation and maintenance of a GIS road network with travel time attributes is resource-intensive. Proprietary New Zealand road network data with associated travel time estimates exist, but these are relatively expensive, which puts them beyond the means of many researchers and organisations. Building on the work of Brabyn and Skelly (2002), this paper discusses the methods used to produce an open-source road network analysis dataset with travel time as a resistance attribute. The intention is to produce a publicly accessible network analysis dataset suitable for modelling relationships relevant to health geography, and that can be updated relatively efficiently.