Validating social media data reliability for cultural ecosystem services mapping in data-scarce GIAHS: A mixed-methods approach in hani rice Terraces

Cultural ecosystem services (CESs) are subjective and intangible, making their assessment challenging. While the application of social media data in assessing and mapping the cultural services of the ecosystem has made rapid progress, this approach relies on active social media users and large sample size of social media data. The reliability of social media data-based method in less developed areas, where such conditions may not be met, remains largely unverified. This study focuses on the Yuanyang Hani Terraces, a Globally Important Agricultural Heritage site located in the remote mountains of southwest China. We compared the value and distribution of CESs as revealed by social media data and questionnaire data, and estimated the applicability of social media data in regions with limited data availability. The results show that: (1) Among the places identified as possessing CESs via questionnaire method, 90 % of places with aesthetic value (AV), 90 % of places with cultural heritage value (CHV), 91 % places with cultural diversity value (CDV), and 80 % of places with scientific & educational value (SEV) were also identified using social media data. The intraclass correlation efficient value for the two-method reached 0.96,0.84,0.79, and 0.76, respectively, indicating a high consistency level. Furthermore, social media data-based method identified more CES places than the traditional method. (2) the CESs in the core area of Hani Terraces showed a pattern of AV > SEV and CHV > CDV. Places with AV mainly located in Duoyishu-Aichun area, Bada-Laoyingzui area and Laohuzui aera. Places with CHV included Shengcun (a local town), museum, famous folk villages, and county houses cluster. Places with SEV were mainly two traditional villages, the museum, and the tourist center where documentary and exhibitions of the Hani Terraces were provided. This study showed that social media data provides highly consistent results with traditional questionnaire method for assessing and mapping CES and presents a cost-effective alternative for similar studies in less developed areas where social media data is sparse.