MULTISCALE DATA MODELS FOR SPATIAL ANALYSIS, WITH APPLICATIONS TO MULTIFRACTAL PHENOMENA

Most of the discussion in the CIS community is concerned at the highest level with the support of managerial issues and at the lowest level with engineering considerations. Scientific considerations of analytical flexibility and conceptual suitability tend to be slighted. One reason for this situation is the complex, multiscale, and "heterodimensional " world with which the scientist is concerned. The new models of spatial fractals and temporal chaos are making aggressive inroads in our understanding of complex systems, and they deserve to inform considerations of the data models that will fruitfully capture variation in space, time, and scale. I present the outline of a data model that I have found useful, along with examples of its use. The model is well suited to current advances in data organization, hardware, and parallel algorithms