Visual Data Exploration and Analysis - Report on the Visualization Breakout Session of the SCaLeS Workshop

Bethel, E. Wes ; Frank, Randy ; Fulcomer, Sam ; Hansen, Chuck ; Joy, Ken ; Kohl, Jim ; Middleton, Don

Lawrence Berkeley National Laboratory

关键词: 99 General And Miscellaneous//Mathematics, Computing, And Information Science;    Computers;    Implementation;    Greenhouse Gases;    Exploration;   

DOI  :  10.2172/815467
RP-ID  :  LBNL/PUB--886
RP-ID  :  AC03-76SF00098
RP-ID  :  815467

美国|英语

来源: UNT Digital Library

【 摘 要 】

Scientific visualization is the transformation of abstract information into images, and it plays an integral role in the scientific process by facilitating insight into observed or simulated phenomena. Visualization as a discipline spans many research areas from computer science, cognitive psychology and even art. Yet the most successful visualization applications are created when close synergistic interactions with domain scientists are part of the algorithmic design and implementation process, leading to visual representations with clear scientific meaning. Visualization is used to explore, to debug, to gain understanding, and as an analysis tool. Visualization is literally everywhere--images are present in this report, on television, on the web, in books and magazines--the common theme is the ability to present information visually that is rapidly assimilated by human observers, and transformed into understanding or insight. As an indispensable part a modern science laboratory, visualization is akin to the biologist's microscope or the electrical engineer's oscilloscope. Whereas the microscope is limited to small specimens or use of optics to focus light, the power of scientific visualization is virtually limitless: visualization provides the means to examine data that can be at galactic or atomic scales, or at any size in between. Unlike the traditional scientific tools for visual inspection, visualization offers the means to ''see the unseeable.'' Trends in demographics or changes in levels of atmospheric CO{sub 2} as a function of greenhouse gas emissions are familiar examples of such unseeable phenomena. Over time, visualization techniques evolve in response to scientific need. Each scientific discipline has its ''own language,'' verbal and visual, used for communication. The visual language for depicting electrical circuits is much different than the visual language for depicting theoretical molecules or trends in the stock market. There is no ''one visualization too'' that can serve as a panacea for all science disciplines. Instead, visualization researchers work hand in hand with domain scientists as part of the scientific research process to define, create, adapt and refine software that ''speaks the visual language'' of each scientific domain.