Abstract
Weather visualization is critical for operational forecasters and weather researchers to analyze, monitor, and predict severe weather events such as storms, tornadoes, and hurricanes. Usually measured weather datasets such as NEXRAD Level II radar data contain some errors and have temporal aliasing issues, so that the raw data cannot represent the true meteorological patterns. Existing weather visualization packages usually ignore these problems, thus they fail to unveil some detailed information and cannot meet the needs of professional meteorologists. Most displays generated by those packages are limited to 2D images, 3D point clouds, or iso-surfaces. Those displays are not able to accurately represent the details of the data. The limitation of visualizing data from a single radar site at one single time step in most packages blocks desirable information in the datasets.
We developed an efficient and accurate visualization tool capable of displaying the long-track 3D NEXRAD Level II volumetric weather data from multiple sites. The data were captured and integrated from three sites (KEAX, KILX and KSLX), and rendered by using the hardware-accelerated volume-rendering technique. Customized transfer functions are also provided for users to map data values to optical properties such as color and opacity to create different views. We also implemented a compression algorithm to compress 3D textures in order to largely reduce the consumption of hard disk space and to enhance the rendering speed without degrading image quality. An example study shows the tracking of a 24-hour supercell storm observed on March 12, 2006, in the Midwest of the United States. Images and animations from our implementation coupled with results from a set of experiments demonstrate that our methods and approaches are fast and robust, making them suitable for processing and rendering huge amounts of Doppler Level II weather data from multiple stations.
Date of this Version
12-5-2007
Department
College of Technology
Department Head
Gary R. Bertoline
Month of Graduation
December
Year of Graduation
2007
Degree
Master of Science
Advisor 1 or Chair of Committee
Gary R. Bertoline
Committee Member 1
Bedrich Benes
Committee Member 2
Laura Arns