Modeling Recreation on Public Lands Using Social Media and Crowd-Sourced Data

• Activity Number: 408 - Modeling Visitation of Public Lands and Analyzing Visitors’ Sentiments Using Social Media
• Type: Topic-Contributed
• Date/Time: Thursday, August 12, 2021 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Samantha Winder* and Emmi Lia and Lesley Miller and Spencer Wood
• Companies: University of Washington and University of Washington and University of Washington and eScience Institute, University of Washington
• Keywords: Outdoor recreation; Social media; Volunteered Geographic Information; Community science

Abstract

Outdoor recreation provides countless individual and societal benefits, yet public land managers often lack fine-scale data on the amount and character of recreational use. Volunteered geographic information (VGI) data have the potential to fill some of these gaps, since they are highly correlated with on-site measures of visitation. Together with information about local weather and calendar variables, VGI can be used to estimate the absolute number of weekly visitors to individual trails. We present a model of public land visitation based on content posted to three social media platforms (Flickr, Instagram, Twitter) and two online trip report platforms (AllTrails, Washington Trails Association). The model is parameterized using on-site data collected at 27 sites in Western Washington between 2016 and 2020 (more than 10,000 site-days of visitation counts), supplemented by crowd-sourced estimates from visitors who count the number of vehicles that are parked at trailheads. We present results from testing this visitation model at out-of-sample trail sites across the United States, and discuss its potential for informing land management decisions.