HR: 0800h
AN: IN31A-1138    [Abstracts]
TI: Data Mining Architecture For The Analysis Of Remote Sensing Data
AU: * Mehra, V
EM: vmehra2@uiuc.edu
AF: Department of Civil and Environmental Engineering, 205 N. Mathews Avenue University of Illinois, Urbana, IL 61801 United States
AU: Kumar, P
EM: kumar1@uiuc.edu
AF: Department of Civil and Environmental Engineering, 205 N. Mathews Avenue University of Illinois, Urbana, IL 61801 United States
AU: Bajcsy, P
EM: pbajcsy@ncsa.uiuc.edu
AF: National Center for Supercomputing Applications, 1205 W. Clark Street University of Illinois, Urbana, IL 61801 United States
AU: Feng, W W
EM: wfeng2@uiuc.edu
AF: National Center for Supercomputing Applications, 1205 W. Clark Street University of Illinois, Urbana, IL 61801 United States
AU: Clutter, D
EM: clutter@ncsa.uiuc.edu
AF: National Center for Supercomputing Applications, 1205 W. Clark Street University of Illinois, Urbana, IL 61801 United States
AU: Sinha, P
EM: psinha3@uiuc.edu
AF: Department of Civil and Environmental Engineering, 205 N. Mathews Avenue University of Illinois, Urbana, IL 61801 United States
AU: Tcheng, D
EM: dtcheng@ncsa.uiuc.edu
AF: National Center for Supercomputing Applications, 1205 W. Clark Street University of Illinois, Urbana, IL 61801 United States
AU: White, A B
EM: abwhite@uiuc.edu
AF: Department of Civil and Environmental Engineering, 205 N. Mathews Avenue University of Illinois, Urbana, IL 61801 United States
AB: Large volumes of data have been collected for hydroclimatological studies through satellites such as TERRA and AQUA and numerous other sources. However, the volume of available data has far outstretched our ability to effectively use them for hypothesis testing, modeling and prediction. Traditional tools of scientific inquiry such as statistical analysis and data assimilation have several limitations in dealing with such large volumes. With the recent emergence of the field of data mining, there is an opportunity for exploring these datasets through systems that can handle large data sets through data assembly, preprocessing and integration tasks along with providing a range of data mining functionality. A system to achieve this objective is developed based on the integration of I2K (Image to Knowledge) and D2K (Data to Knowledge) software developed by NCSA (National Center for Supercomputing Applications) along with ArcEngine. Our System has the capacity to access very large multivariate datasets; represent heterogeneous data types; integrate multiple GIS data sets stored across many GIS file formats; analyze variable relationships and model their dependencies; and visualize input data, as well, as extract features, integrate data sets and data mining results. Data mining applications to scientific data sets at the continental to global scales will enable us to better parameterize the various natural processes for weather and climate models and thereby improve their predictability.
DE: 0300 ATMOSPHERIC COMPOSITION AND STRUCTURE
DE: 3200 MATHEMATICAL GEOPHYSICS (0500, 4400, 7833)
DE: 3300 ATMOSPHERIC PROCESSES
DE: 4200 OCEANOGRAPHY: GENERAL
DE: 7200 SEISMOLOGY
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005