An object-oriented framework for distributed hydrologic and geomorphic modeling using triangulated irregular networks

We describe a new set of data structures and algorithms for dynamic terrain modeling using a triangulated irregular network (TINs). The framework prov ides an efficient method for storing, accessing, and updating a Delaunay tr iangulation and its associated Voronoi diagram. The basic data structure co nsists of three interconnected data objects: triangles, nodes, and directed edges. Encapsulating each of these geometric elements within a data object makes it possible to essentially decouple the TIN representation from the modeling applications that make use of it. Both the triangulation and its c orresponding Voronoi diagram can be rapidly retrieved or updated, making th ese methods well suited to adaptive remeshing schemes. We develop a set of algorithms for defining drainage networks and identifying closed depression s (e.g., lakes) for hydrologic and geomorphic modeling applications. We als o outline simple numerical algorithms for solving network routing and 2D tr ansport equations within the TIN framework. The methods are illustrated wit h two example applications, a landscape evolution model and a distributed r ainfall-runoff model. (C) 2001 Elsevier Science Ltd. All rights reserved.