Evaluating the quality of VGI and authoritative data in red-crowned crane conservation: a comparative study

Volunteered Geographic Information (VGI) has emerged as a valuable resource in citizen science, yet concerns about its data quality remain. While prior studies have focused on specific data types or platforms, few have systematically evaluated heterogeneous VGI datasets within a unified application context. Addressing this gap, this study evaluates the data quality of VGI by comparing vector data from eBird and raster data from OpenStreetMap against authoritative datasets from GBIF and CASEarth, focusing on red-crowned crane habitats in Hokkaido, Japan. The evaluation focuses on thematic accuracy, coverage, and hotspot distribution, which are highly relevant to ecological conservation. For vector data in species distribution, VGI shows higher thematic accuracy and broader spatial coverage, while authoritative data spans a longer time range. Hotspot analysis reveals a more balanced distribution in VGI and stronger clustering in authoritative data. For raster data in land use, OSM performs well in forest classification but poorly in cropland and grassland, with notable coverage gaps. Additionally, hotspot analysis reveals differing spatial clustering patterns between VGI and authoritative datasets. These findings underscore the varying reliability of VGI across data types and sources, and call for tailored validation strategies to enhance its utility in ecological research.