HR: 11:15h
AN: SM52A-04 INVITED [Abstracts]
TI: Next Generation Data Mining and Visualization Techniques: Application to Global Hybrid
Simulations
AU: * Omidi, N
EM: omidi@scibernet.com
AF: SciberNet Inc., 777 Pacific Coast Highway, Solana Beach, 92075
United States
AU: Karimabadi, H
EM: homak@scibernet.com
AF: SciberNet Inc., 777 Pacific Coast Highway, Solana Beach, 92075
United States
AU: Scudder, J
EM: jds@space-theory.physics.uiowa.edu
AF: University of Iowa, Univ Iowa
Dept Physics & Astronomy
Rm 203 Van Allen Hall, Iowa City, 52240
United States
AU: Driscoll, J
EM: driscol@scibernet.com
AF: SciberNet Inc., 777 Pacific Coast Highway, Solana Beach, 92075
United States
AU: White, H
EM: hal.white@bwanalytics.com
AF: BW Analytics Inc., 12340 El Camino Real
Suite 350, San Diego, 92130
United States
AB:
Global hybrid simulations, in which electrons are treated as a fluid and ions kinetically, provide high resolution "images"
of the magnetosphere on ion temporal and spatial scales. Visualization and analysis of data from these simulations pose
serious challenges that need to be addressed. For example, a typical global 2-D simulation run generates over 100 Gb of data
even when individual particle data are saved much less frequently than moments and field data. A 3-D global run easily
generates over TB of data. Due to their large size, these data sets are often stored in segments over distributed platforms.
An added complication is that no off-the-shelf software exists and packages such as IDL, OpenDx and VTK require significant
customization. In this talk, we provide an overview of our customized software and illustrate their usage through examples.
In particular, three aspects of data analysis techniques are described in some detail. One, has to do with tools developed
for visualization of data in various perspectives and application of standard analysis techniques to them. The other,
involves the use of intelligent software for data mining and event identification purposes. Third, is the development of
special software to facilitate comparison between simulation and spacecraft data. In all cases, we place our focus on bow
shock or magnetopause related processes.
DE: 2724 Magnetopause, cusp, and boundary layers
DE: 2728 Magnetosheath
DE: 2753 Numerical modeling
SC: SPA-Magnetospheric Physics [SM]
MN: 2004 AGU Fall Meeting