ITR: Gleaning Insight into Large Time-Varying Scientific and Engineering Data

University Of California-Davis, Davis CA

Investigators

Abstract

The project consists of a basic research component and a significant system/tool building component such that our research and development can be fully utilized and extensively evaluated by the application scientists with which we will closely collaborate. Such efforts can only be made possible with a four-year ITR project at the proposed scale. In addition to the large number of applications that we will have access to, the proposed research is particularly unique and important in the following ways: (i) Developing simulation-time data reduction (i.e., encoding and feature extraction) techniques. Addressing the visualization challenges early in the visualization pipeline is a largely ignored direction in the past. (ii) Investigating flexible modeling of the extracted temporal and spatial features to facilitate efficient rendering and manipulation at different levels of abstraction. (iii) Designing highly optimized rendering methods, coupled with novel interaction techniques, to facilitate interactive browsing and exploration of the encoded data or extracted features directly. (iv) Seeking affordable solutions based on commodity-hardware accelerated, progressive visualization to allow scientists to conduct as much of the iterative part of the visualization and data understanding task as possible on their desktop PC. (v) Placing a strong emphasis on realizing software tools and delivering them to our application scientist collaborators as well as to the broader scientific community. These tools will be usable for the scientist's routine work, not just for proof of concept. Complementing the research and development components of this project is an educational component, directed at both graduate and undergraduate students, aimed at inspiring interest and fostering creativity through hands-on experiences. We hope to foster greater interaction and cross-fertilization of ideas between the inter-related disciplines of visualization, scientific computing, and data analysis.