The Identification of Data Primitives to Separate the Concepts and Semantics of Land Cover and Land Use: the Example of ‘forest’

This paper is concerned with data and classifications that confuse the concepts of land cover and land use. Conceptual confusion is problematic for data integration and modelling. This has resulted in calls for the separation of land use and land cover from the global land monitoring community. Text mining approaches are combined with social network analyses as a method for unravelling the different concepts embedded in land cover and land use semantics and applied to descriptions of forest cover and use. Whilst the results show the distinct biological dimension to land cover descriptions and the socio-economic character of land use, they reveal the deep degree of semantic confusion embedded in land cover and land use descriptions. The implications for this lack of internal semantic accuracy and consistency in land resource inventories are discussed and the case made for separating the concepts of land cover from land use.