Analyzing national parks visitor activities using geotagged social media photos

Visitor activities within national parks are important information for park management. Currently, national parks and nature-based recreation area worldwide have no systematic visitor activity data collection schemes. Traditional visitor movement or activity data collection methods used in tourism studies are often costly in money and time and difficult to implement. This paper presents a framework and formal procedure that uses geotagged photos crowdsourced from social media platforms to uncover visitor activities within national parks. First, the daily individual space-time paths are constructed from the photo geotags and timestamps for all vis-itors. Then the following information is derived from the space-time paths: 1) four activity types including stationary, walking, vehicle riding, and flying; 2) the total time of various visitor activities in a park; 3) visitor flow directions aggregated to a grid; 4) a graph constructed from space-time paths and the directed and undi-rected connections (flows) between POIs; and 5) the centrality of POIs derived from the graph. A case study of four highly visited U.S. national park units from 2008 to 2021 demonstrates the process and capabilities of the method. It also reveals a trend of increased walking time and decreased stationary time in the case study parks over the years. Although this study is only based on Flickr photos, the method can be implemented with data derived from any social media platforms as long as geotags and timestamps are available. The method may be used to monitor visitor activities and movements within national parks and other outdoor recreation areas, especially remote and unmonitored locations.