Authors: Adam Berland*, Ball State University, Jess Vogt, DePaul University, Lara Roman, USDA Forest Service
Topics: Geographic Information Science and Systems, Environment, Natural Resources
Keywords: citizen science, photograph interpretation, street-level imagery, urban forestry
Session Type: Poster
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded
Cities increasingly rely on field inventory data to manage public trees along streets, but field inventories are expensive and time-consuming. To reduce costs, some cities rely on citizen scientists to generate data regarding street tree locations, species, sizes, and health. In this study we leverage citizen scientists to generate data about street trees, but rather than visiting the trees in the field, volunteers use freely and publicly available Google Street View imagery to conduct virtual surveys of street trees. The goals of this research are to first understand the general level of data quality that can be expected from virtual surveys of street trees. To accomplish this goal, we compare virtual survey data to field data collected along the same streets in suburban Chicago. We also track time taken to conduct the work to quantify tradeoffs in data quality and efficiency associated with virtual surveys vs. field surveys. Second, we study how data quality varies according to the analyst’s expertise in conducting street tree inventories. By highlighting tradeoffs in data quality and efficiency across strata of analyst expertise, we provide guidance for communities exploring the application of Google Street View imagery for virtual surveys of street trees. We also outline a workflow for integrating multiple technologies including mobile GIS apps for field data collection, collaborative online platforms for data collection, and desktop GIS for data collation and analysis.