In the study, the researchers used photos that were shared on the photo-sharing service Flickr and tied to a geographic location. The viewshed, or visible area, of each photo was assessed using Geographic Information System software. They identified locations where it appeared that many photos had been taken, and considered these as sites with high landscape “demand,” which were then evaluated and given numerical scores. Next, the team used a modeling method called Maxent to identify locations that appeared to have similar environmental conditions as the locations with high landscape demand. They then evaluated the potential values of these locations, which they called the “potential supply.”
The left map shows landscape demand, with high demand areas indicated in red. On the right map, which indicates potential supply, green and yellow areas are those with similar environmental attributes to locations with high landscape demand, such as proximity to a volcano or open, scenic place. (Yoshimura N. and Hiura T., Ecosystem Services, February 28, 2017)
This method was applied to locations across Hokkaido in Japan, which indicated that landscape demand is higher in national parks followed by areas of farmland and cities. Comparisons between landscape demand and potential supply showed that, for example, Shikotsu-Toya National Park and Daisetsuzan National Park have greater potential supply than demand, while Kushiro Shitsugen National Park has greater demand than potential supply. “These kinds of comparisons could enable investigations into whether locations with high landscape demand should be targeted for environmental preservation or whether farmland could be utilized as a tourism resource,” says Tsutom Hiura. The researchers plan to further improve the method and test it with other social networks.
