Using geo-data and social media images to explore the supply and demand of cultural ecosystem services for terraces in China

Terraces are a unique type of agro-ecosystem that are vital for regional food security, biodiversity, and the provision of cultural ecosystem services (CES) to society. This study introduces a novel approach to map the supply–demand balance of terrace CES (TCES) in China by integrating geo-data and social media images. Firstly, a TCES supply assessment framework is developed, comprising scenic attractiveness and heritage attractiveness, to assess TCES supply. Secondly, based on 55,616 geotagged Weibo images, the EfficientNet model classifies images into seven categories, with a questionnaire linking these to four CES types: aesthetic services, heritage & cultural services, recreation & tourism services, and spiritual & emotional services. Population data is also used to assess TCES demand. Finally, the supply–demand ratio and bivariate Moran’s I examine the balance and spatial autocorrelation of TCES. The results show: 1) High supply areas are mainly in southern China, while demand is more scattered in several hotspots; 2) There is a significant positive spatial autocorrelation between supply and demand, where higher supply promotes greater demand. High CES supply and demand cluster in Zhejiang-Fujian Hills and Yunnan-Guizhou Plateau, with imbalances occurring in metropolitan areas or mountainous areas along provincial boundaries. These findings and methodologies provide valuable insights for the planning and management of terraces in China, as well as for future CES-related studies.