Of central importance to hydrologic and soil erosion modelling is the development of an adequate representation of terrain. Such a representation must be capable of yielding precise estimates of elevation, slope, and aspect at any point. Little progress has been made in solving this problem because of its complexity, however. Present models for predicting effects of land use practices and agricultural management practices on nonpoint source pollution, for example, are limited by their inability to adequately consider three-dimensional effects. Based on the idea that terrain can be represented in the form of discrete polygons, each bounded by two contour segments and two interpolated streamlines, the investigators will develop a technology to handle spatial variability of the attributes in large watersheds and river basins. This project will refine and begin testing of an established hydrological model that integrates algorithms developed for more specific models. This innovative study aims to develop methods that encompass three-dimensional effects and that are more tractable, accurate, and practical for use by engineers, soil scientists, and other practitioners than are current methods. Explanation of means of representing terrain, and the central statistical questions that must be dealt with in working out this representation, will contribute to the development of geographic information systems. These methods have the potential to facilitate progress toward modelling the effects of management and land use practices on nonpoint source water quality and quantity. The techniques developed may also be applicable to subsurface water movement and to related groundwater applications, which in turn are likely to be essential considerations in the development of reliable remedial measures.
