Geolocated social media data for measuring park visitation in Shenzhen, China

Abstract

Understanding park visitation patterns and factors that correlate with park use is conducive to urban green space management and planning. Although a growing number of studies have indicated geolocated social media can be used as a proxy for recreational use analysis, most of them relied on a single or similar social media platform(s). In this study, we used geolocated social media data among four popular platforms in China to estimate park visitation and explore the influence of 13 potential factors on park usage in Shenzhen by the geographical detector model. A park visitation index was introduced to estimate park usage, and it indicated large parks, such as natural and city parks, have a higher park visitation index than community parks in Shenzhen. In contrast to the check-in data, the real-time user data has a higher potential to describe park visitation citywide. Our findings demonstrate park size, sports and recreation amenities, the length of trails, and the nearby building density of a given park are likely to influence park visitation. The enhanced interactive relationship of 13 potential determinants of park usage can provide implications for urban park management and planning.