Landscape values indicate how humans perceive and evaluate the landscape. In our study areas, two hydroelectric proposals have the potential to alter the landscape dramatically, particularly the river (reservoir) and riparian land. An understanding of the spatial patterns of landscape values, especially the social and cultural values which are intangible and underestimated in energy planning processes, can help decision makers to anticipate public concerns and adjust or abandon project proposals accordingly. Intangible landscape values can be revealed in part by leveraging social media. Such data sources have two benefits in relation to the challenges of previous, manual approaches: they give us access to people (e.g. youth) who are often absent in conventional participation methods, and provide large datasets at low cost. We collected photos and captions that were geo-tagged to the study areas on the social media site Instagram, and built a filtering model to increase validity of data for the calculation of point density (specifically, kernel density estimation). The density maps reveal that: (1) landscape values vary over space; (2) aesthetic value was most widespread (not surprising given typical uses of the Instagram platform); (3) town areas, especially the old ones, and popular viewpoints were most likely to be attractors for multiple values. People tend to accept and appreciate familiar landscapes, thus proponents should make particular allowances for locations of key values and multiplex values.
