ABSTRACT
Fields have been considered in Chapter 2 as being important models for spatial data, as they are widely recognised as being flexible and powerful for capturing spatial variability and heterogeneity in geographical reality, and are thus preferred for purposes of uncertainty handling (Goodchild, 1989; Kemp, 1997; Vckovski, 1998). In the field domain, variables may be discrete-or continuous-valued. While both categorical and continuous variables are relevant to fields, the latter seem be closer to the classical notion of continuous variation, and often serve to bridge the gap between idealised discreteness and realistic continua. In light of this, it is sensible to start the exploration of geographical uncertainties from the domain of continuous fields. This chapter aims to provide an account of fieldbased methodologies for handling spatial uncertainties, backed up by experimentation with real data.
