Use of iNaturalist Biodiversity Contribution Data for Modelling Travel Distances to Parks Across the United States

Abstract. Crowdsourcing platforms have become an important data source for modelling and observing human behavioural and social activities, such as mobility, social interactions and urban dynamics. This study uses observation data from iNaturalist, an online social network of voluntary users sharing biodiversity information, which was collected from 20,434 parks in the United States. It explores the relationship between park characteristics and the mean travel distance of users to parks. The latter is based on the average of distances between an iNaturalist user’s typical main area of iNaturalist contributions and the locations of the user’s observations falling inside a park of interest. The DBSCAN clustering algorithm is used to determine each user’s main contribution area. An Eigenvector Spatial Filtering (ESF) model shows that the log of the average distance travelled to parks is positively associated with certain park management types (e.g. National Parks, State Parks) and biodiversity, but negatively associated with the population around a park. The results provide insights into the nature of iNaturalist user visitation patterns to parks which can be used for targeted outreach campaigns and a more user-centric approach to promote park attractions and biodiversity conservation.