Dynamic Modeling, Geostatistics, and Fuzzy Classificaiton: New Sneakers for a New Geography?

O lId Methodological Sneakers: Fashion and Function in a CrossTraining Era (Bauer et al. 1999) summarizes the recent focus and shortcomings of physical geography and suggests why the conventional approach to the discipline may be insufficient to compete successfully in a future dominated by multidisciplinary teams from other earth and biological sciences, working to solve real-world problems. The synthesis offered by Bauer, Veblen, and Winkler in their Forum introductory essay (pp. 679-87) may be just as apposite for recent work in pedology as in other subfields of our discipline. In recent decades, soil geographers and others have investigated soils at scales ranging from the molecular to the pedon (plot), polypedon (field), watershed, county, region, country, and globe. Much of this research, however, has been a static inventory and analysis of individual pedons or hillslopes that has avoided spatial and temporal interactions of interest at the management scale of farm field or catchment (Usery et al. 1995; Bouma 1997). Work at these scales has, until recently, focused on the production of maps and related database products (e.g., the SSURGO, STATSGO and NATSGO databases, Reybold and TeSelle 1989; USDA-NRCS 1991; Digital Soil Map of the World, FAO 1974-1978). The advent of Geographic Information Systems (GIS) and related geospatial technologies, while greatly increasing the availability and use of these products (Wilson 1999a), has also served to reinforce the conventional paradigms of space on which they depend. In reaction, an increasing number of studies has critically examined the methods used to construct these products and how these methods have constrained the physical geographer in his/her ability to convey understanding about operation of the landscape in response to natural and anthropogenic forces (e.g., Baker and Gersmehl 1991; Wilson et al. 1996; Burrough et al. 1997; Zhu 1997). These investigations have indicated that there are several important new opportunities that can be used to: (1) extend the discussion of Bauer et al., and (2) show that there are several useful extensions to the conventional spatial and temporal paradigms that will enable the multidisciplinary real-world problem solving that is likely to dominate the academic enterprise of the twenty-first century. Although computer technology plays an important role in the enablement of these new methods, the real innovation involves a fundamental change in the way space, objects in space, and space-time interactions are perceived, described, and recorded. These ideas are discussed here with reference to recent work in physical geography and soil science.