Vagueness and the trade-off between the classification and delineation of geographic regions – an ontological analysis

Abstract

For inherently vague and granular phenomena such as ecoregions, ecosystems, biomes, and biotopes, the interplay of granularity and vagueness leads to a trade-off in the classification and delineation of such phenomena: the goal of preciseness (lack of vagueness) of the delineation contradicts the goal of building a sophisticated classification system using the Aristotelian method of classification. This trade-off is based on the reliance on local qualities for a precise delineation of particular regions and the reliance on nonlocal qualities that serve as differentia in the Aristotelian classification. An ontological analysis of the logical interrelations between vagueness, granularity, and scale is critical for developing logically rigorous, nonlocal, and nonarbitrary classification and delineation systems for inherently vague and granular geographic phenomena.

On the Integration of Regional Classification and Delineation Systems into The National Map

“…To capture the granular and scale-dependent ways in which quality patterns characterize geographic regions, the technical notion of homogeneity (Bittner 2009) is used (see Table 1.) Consider, for example, the geographic regions Central Great Plains (CGP) and Flint Hills (FH), labelled respectively ''9.4.2'' and ''9.4.4'' in Figure 1a.…”

“…To integrate the three systems, shared structural properties are formalized using the top-level ontology of (Bittner, Donnelly, and Smith 2009;Bittner 2009 …”

“…Because of the fundamentally vague and granular nature of many of the qualities that characterize geographic regions (Mark and Csillag 1989;Ahlqvist 2005;Waterton 2002;Bailey 2004;Omernik 2004;Kronenfeld 2003), there is no easy way to overcome the trade-off between the local character of the precise quality-based delineation of geographic regions and the non-local character of the quality-based classification of geographic regions (Bittner 2009). In what follows, one possibility to overcome this trade-off is discussed.…”

“…The resulting system inherits the best aspects of all three of the original systems and could be integrated in TNM to meet the needs of all TNM users. A previous ontology-based analysis of quality-based classification and delineation systems for geographic regions (Bittner 2009) provides the theoretical framework for this article. It also provides the basis for a symbolic, logicbased representation that may facilitate the integration of TNM into the Geo-semantic Web (Semantic Web 2009).…”

“…By contrast, quality-based classification systems for geographic regions rely on nonlocal combinations or patterns of qualities that differentiate classes (or universals) independently of the specific location of their instances. In this article, based on work reported elsewhere (Bittner 2009), the eco-region classification and delineation systems of the US Forest Service (USFS 1995) and the US Environment Protection Agency (EPA 1997) are used to illustrate the trade-off between the local character of the precise quality-based delineation of geographic regions and the non-local character of their quality-based classification. Consider the map displayed in Figure 1a.…”