cient Visualization of Large { Scale Data onHierarchical

A multi-resolution approach is presented for data on a large class of hierarchical and nested grids. It is based on a procedural interface and a set of hierarchical and adaptive visualization methods. Such a method consists of a recursive traversal of mesh elements from the grid hierarchy combined with an adaptive stopping according to some error indicator which is closely related to the visual impression of data smoothness. During this traversal user data is only temporarily and locally addressed on single elements. No in advance mapping onto prescribed formats is necessary. The user only has to supply a set of element access routines as an interface to his speciic data structures. As no extra storage is required, also large, economically stored computational grids can be handled on workstations with moderate local memory. Signiicant examples illustrate the applicability and eeciency on diierent types of meshes.