Authors: Jeffrey Olson*, University of Wisconsin - Whitewater, Derek van Berkel, School for Environment and Sustainability, University of Michigan
Topics: Rural Geography, Geographic Information Science and Systems, Economic Geography
Keywords: Rural, Natural Amenities, Social Media, machine learning, GIS
Session Type: Paper
Presentation File: No File Uploaded
Rural areas in the United States serve many economic, social, and recreational functions that are often overlooked or simplified in everyday discourse. This is partially due to the economic and media dominance of large cities, but may also be due to lack of spatially refined data representing rural spaces. Research addressing changes in rural areas have tended to rely on county-level data representing population and employment growth and decline; often seeking to find relationships with urban proximity and the endowment of natural amenities. Amenities, as measured by the presence of water bodies, pleasant climates, and varied landscapes among other measures, have been found to explain a significant portion of rural growth since World War II. Less explored is how people record and consume these pleasant rural environments as amenities. The explosion in geographically-referenced social media data may yield new insights as to what spaces people perceive as being beautiful, evidenced by shared photographs that focus on rural scenery. In this talk, we use machine learning image processing techniques to identify nature photographs geotagged by individuals in rural areas. We then use clustering algorithms to classify rural spaces based on their social media, amenity, and socioeconomic characteristics. We identify different types of rural spaces at a 10 kilometer resolution in an effort to find commonalities in places of similar growth trajectories.