Most of the data that we have used are available from official bodies and Google Earth. The novelty here lies in the fact that the database we produced was georeferenced in order to create a GIS with different theme-based layers which we then used to analyse the whole dataset. We began by creating a layer that included both new HSR lines and railway stations currently in service.3 We then added a layer containing agglomerations with over 10,000 inhabitants and selected those located within 25 km of a railway station: the catchment area. This GIS made it possible to calculate both the distances between different railway stations and also those from them to the nearest population nuclei. This was particularly useful for identifying cases in which a station was located on the outskirts of, or at some distance from, its nearest population nucleus. Finally, we added a third layer corresponding to the region in which each station was located. This allowed us to associate each station with the per capita GDP of the area to which it belonged. We used all this information due to its relevance and because it corresponded to recent and comparable data (2007–2014) which were available for all the 174 HSR stations considered in this study. The next question is how we used this information and what types of challenge we encountered. Having information relating to the real number of HST users, by station, would be ideal for checking our findings. Unfortunately rail companies no longer provide this kind of data.
There are four previous steps to this methodology: (1) assigning data relating to population; (2) locating railway stations with respect to their corresponding urban nuclei; (3) calculating distances between stations; and (4) assigning the regional GDP parameter.