Assessing the potential of social media for estimating recreational use of urban and peri-urban forests

Abstract Social forest functions including recreation are important for increasingly urbanised societies. For effective management of forest recreation areas, monitoring visitor frequencies is crucial. Increasingly, attempts are being made to incorporate recreational use data into National Forest Inventories (NFI), but given the large scale of national assessments, such data is often elusive. In this study we explore the potential of geotagged social media data for assessing visitor frequencies and explore recreational activities through text-based social media data. We analysed data from Twitter, Flickr and Instagram, both at local scale for 10 NFI forest sites, as well as at national scale to assess recreational use. Data availability was significantly correlated between the three platforms, even though absolute counts differed markedly. The model of recreational visitation based on social media data correlated significantly with an existing potential recreational model, indicating that social media data are a valid source of information for recreational use and can be used in future studies to assess recreational potential. Although data availability limits assessments for small areas of forests, large scale assessments using social media are feasible, and provide a potentially more empirically grounded assessment of recreational potential than theoretical models alone. We suggest that future work should aim at integrating social media data into traditional theoretical recreational models as part of a method triangulation, particularly for areas where recreational usage by visitors is high, but population counts are low. However, because social media data are provided by commercial platforms, we believe that more research is needed into harvesting and analysing other forms of content generated by users to decrease the dependency on commercial social media platforms that may or may not be available in the long run, and can be run locally or through central organisations involved in forest and landscape monitoring and observation.