Land Type Categories as a Complement to Land Use and Land Cover Attributes in Landscape Mapping and Monitoring

The purpose of land cover and land use descriptions varies, and this influences how these concepts are perceived in different contexts. The increasing need for spatial data for multipurpose monitoring and modeling also increases the demands for compatibility, repeatability, detail, and well-documented criteria. We suggest that threshold values along a continuous scale can be used to create nominal classes for a common conceptual framework. However, the exact values of these thresholds need to be based on well-defined functional and systematic criteria. Ecological and environmental gradients are often mosaic and complex, and several types of land use may coexist at the same site. In reality, land use can be seen as a “shifting cloud” of activities varying in both time and space. We advocate the use of strict definitions of land cover as physical structures and land use as human activities, which raises the need for a complementary concept, which we call “land type,” with stable threshold values based on mutually exclusive functional criteria. Such functional criteria often put clear limits to what spatial resolution is appropriate, since the suitability for a certain purpose (e.g., agriculture or forestry) is determined by the user of the land, rather than by the independent observer. Our example of land type categories comprises a two-level hierarchical classification with seven main types and altogether 28 subtypes. As an example, we discuss the overlapping Swedish definitions of forest and arable land. The criteria that define our main land types are less dependent on how the area is managed at a specific moment in time, and they are therefore less sensitive to short-term variation. The land types define the limits for what land cover and land use can be expected at a certain site, given, for example, ground conditions, water, or artificial structures. Since such land types need to incorporate functional and qualitative understanding and interpretation, human visual interpretation is needed, whereas automated remote sensing methods are suitable mainly for the structural aspects of land cover.