| The goal of the study is to investigate representations of scale in Earth System Science (ESS), physical geography and related disciplines and to formulate a new geospatial data model on the basis of such investigation. In the context of this research data model is understood as a structure in computer memory that is utilized by a program for persistent storage of geospatial information. A new data model is based on the analysis of conceptualization of scale in ecosystem studies, soil science and related fields of research.; Approach to the analysis of conceptualization employed the methods of formal ontology and was based on mereology (a theory of part-of relations), a theory of classification, and a theory of interconnectedness. Axiomatization of the theories is presented using first order predicate logic. This methodology was applied to three information pools: Classification of Ecological Subsection of Southeast Alaska, National Hierarchical Framework of Ecological Units of the U.S., and United States Department of Agriculture (USDA) Soil Taxonomy.; The study resulted in a new abstract data model that is depicted in a series of Unified Modeling Language (UML) diagrams. The new model is called CIR data model after its three major components: C&barbelow;lasses, I&barbelow;ndividuals and R&barbelow;egions. The primary component of this model are individuals that represent real-world objects or particulars. Individuals can be connected with each other through part-of relation and can form hierarchies. Each individual can instantiate one or more classes. Classes represent universals or categories of real-world objects. Classes form hierarchies through the is a (subclass) relation. Each individual is located in one or more non-overlapping regions. Regions are related to each other through the relationship of connectedness.; The CIR model addresses many limitations of the existing geospatial data models. The main advantage of the new model is its ability to consistently handle multiple scale levels. Also the CIR model allows researchers to approach such typical problems of geospatial data models as reliance on static content, flatness of the cartographic representation and crispiness of boundaries. The model can also be used to facilitate interoperabilily between geospatial datasets and geographic classifications. |
